EMBODIED EXPERIENCES IN IMMERSIVE VIRTUAL ENVIRONMENTS:
EFFECTS ON PRO-ENVIRONMENTAL ATTITUDE AND BEHAVIOR
A DISSERTATION
SUBMITTED TO THE DEPARMENT OF COMMUNICATION
AND THE COMMITTEE ON GRADUATE STUDIES
OF STANFORD UNIVERSITY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
Sun Joo Ahn
May 2011
ABSTRACT
Immersive virtual environments (IVE) allow users experience vivid
sensorimotor stimuli by digitally simulating sensorimotor information. As a result,
users are able to embody experiences by seeing, hearing, and feeling realistic
perceptual cues linked to those experiences. Embodied experiences are defined as
being surrounded by simulated sensorimotor information in mediated environments
that create the sensation of personally undergoing the experience at that moment.
Based on the theoretical framework of embodied cognition, which stipulates a close
connection between sensorimotor experiences of the body and mental schemas, the
current dissertation studies demonstrated that embodied experiences in IVEs are
able to influence attitudes and behaviors in the physical, non-mediated world.
Two studies explored the effect of embodied experiences on proenvironmental attitude and behavior by having participants embody the experience
of cutting down a redwood tree as a result of using non-recycled paper products.
Focus was also placed on investigating individual elements of embodied
experiences in IVEs and the moderating effect of individual differences in the
capacity to feel presence, the perception that a mediated experience is real. In both
studies, actual pro-environmental behavior is observed and compared to self-reports
of attitude and behavioral intention.
Study 1 (N = 47) compared embodying the tree-cutting experience in IVEs
against mental simulation (MS) of the experience after priming participants with
information on using non-recycled paper products and deforestation problems.
Results demonstrated that participants in both experimental conditions experienced
increased pro-environmental self-efficacy, or the belief that their individual actions
could improve the quality of the environment. However, in terms of proiv
environmental behavior measured by observation of actual usage of paper napkins,
participants who embodied the experience in an IVE demonstrated greater paper
conservation compared to those who merely imagined the experience. Analyses also
confirmed that participants in the IVE condition felt greater presence and personal
control during the embodied experience compared to those in the MS condition.
Study 2 (N = 101) manipulated individual elements that could affect how
participants embodied the tree-cutting experience in IVEs. Two independent
variables were manipulated in two levels in the same tree-cutting context from
Study 1 – immersion (i.e., number and degree of sensorimotor information provided
to the user; High vs. Low) and agency of control (i.e., the agent who controls the
movements of the saw; Self-Move vs. Other-Move). Results demonstrated that
similar to Study 1, any form of embodied experiences increased pro-environmental
self-efficacy. However, agency of control influenced how actual pro-environmental
behavior was manifested in terms of paper conservation and this effect was
moderated by individual differences in the capacity to feel presence. Individuals
high in capacity demonstrated greater pro-environmental behavior when allowed to
fully control the saw, whereas individuals low in capacity demonstrated greater proenvironmental behavior when they were allowed to let pre-recorded saw
movements guide their hands while cutting down the tree. Perceptions of presence
and control are also measured and analyzed to gain greater insight to the process of
embodied experiences in IVEs. In sum, the current studies demonstrated that the
vivid sensorimotor experience of seeing, hearing, and feeling a tree being cut down
in IVEs as a result of using non-recycled paper products is powerful enough to
influence actual paper consumption behavior in the physical, non-mediated world.
v
ACKNOWLEDGEMENTS
So many people have touched my life, provided invaluable guidance, and
showered me with an unbelievable amount of support and encouragement during
my time as a doctoral student at Stanford that I almost feel as if this dissertation is a
collaborative effort of all of my colleagues, friends, and family. These were
arguably the most dynamic five years of my life, and as much as I hate the banality
of this overused expression, time seems to have literally flown by, to the extent that
the fact I am getting ready to graduate seems surreal, even. In just five years, I have
managed to find my partner in life, marry him, obtain a doctoral degree, and secure
a job as an assistant professor. Most importantly, as I write this acknowledgment
while looking back at these emotionally-turbulent turning points of life, my husband
and I are nervously and excitedly anticipating the very imminent arrival of our first
son, Ryan Chu.
It is obvious that graduation will also mean the beginning of topsy-turvy
changes in my life as my new roles as an independent scholar and a new mommy
commence. But before anything else, I think I should take a short moment to
express heartfelt thanks and gratitude to the countless people who helped me grow
both intellectually and emotionally through these events.
Here at Stanford, I would like to thank my advisor, Dr. Jeremy Bailenson,
for being there for me during every hesitant step I took toward becoming an
independent scholar and researcher. He is the one who introduced me to the world
of virtual reality, and with undying patience, allowed me to develop my own area of
interest and to discover the joys of research. Most importantly, when something
needed to be done as soon as possible; when I had a burning question; when I
desperately needed advice – I knew I could depend on him without a shred of doubt.
vi
I cannot even begin to describe what a luxury this was during times of stress and
uncertainty. I probably could not have come so far and grown so much in such a
short span of time without his attention and guidance.
I would also like to thank Dr. Cliff Nass for his ability to see the best in
people. He always knew the right thing to say when I was just about ready to give
up. Anyone can see that he genuinely cares about the welfare of students and his
words of wisdom have a way of resonating in one’s heart for quite a long while. Dr.
Byron Reeves must also be thanked for always knowing the “right” answer and for
having strong faith in my work. He always found time to be a mentor in times of
need despite his incredibly busy schedule and the advice coming from his long
years of experience and expertise was invaluable. Special thanks to Dr. Christian
Wheeler and Dr. Baba Shiv of the Graduate School of Business for reading this
dissertation and providing insightful comments for improvement – the dissertation
has now become a significantly stronger piece of work, thanks to their constructive
inputs. Finally, I would like to thank Dr. Hunter Gehlbach for offering valuable
insights on the theoretical questions leading to this dissertation, helping improve the
actual study design, and providing his questionnaire items. Overall, I was blessed
with incredible opportunities to consult with the best experts that the field has to
offer both in and outside of Stanford and was constantly amazed at their willingness
to help resolve my problems as if those problems were their own.
Furthermore, this dissertation would never have been complete without the
extraordinary support from all the programmers and research assistants of the
Virtual Human Interaction Lab and the VRITS program. Many thanks to Crystal
Nwaneri, Divij Gupta, Felix Chang, Huong Phan, Julio Mojica, Michelle Del
Rosario, Oliver Castaneda, and Jim Zheng for working tirelessly to meet some
vii
unrealistic deadlines for the completion of these studies and always coming to the
rescue in times of dreaded technical failures. Special thanks to Cody Karutz, the lab
manager, who magically made everything happen, no matter how tight the timeline
or how difficult the task.
And thanks to everyone at the front office for dealing with the monstrous
amount of paperwork associated with being a non-resident alien in the U.S. A big
thank you to Susie Ementon who always had answers to every question I had – on
documentation, procedures, life as a grad student, and yes, even on rearing babies. I
would also like to thank all the other staff members, Barbara Kataoka, Mark
DeZutti, Mark Sauer, Katrin Wheeler, and Lisa Suruki, for being helpful in every
possible way.
Outside of Stanford, I must thank Dr. Kyu-ho Youm for his mentorship ever
since I first decided to pursue a doctoral degree and Dr. Eun-ju Lee for her words of
wisdom on various aspects of life, including how to manage a decent work-life
balance as a working mother. I look up to them as my role models both
professionally and personally. Many thanks to all of the professors at the
Department of Communication at Seoul National University, including Dr. Seunggwan Park, Dr. Seung-mok Yang, Dr. Joon-woong Rhee, Dr. Sug-min Youn, and Dr.
In-cheol Choi who helped me set out on my journey at Stanford and believed in my
potentials as an academic from the beginning. A special thanks to everyone at the
Korea Foundation for Advanced Studies who provided me with an unbelievably
generous fellowship for the past five years. It was an incredible luxury to be free
from worrying about tuition, medical insurance, and living expenses for so many
years and it allowed me to focus completely on research.
Next, the reason I am able to look back at the past five years and can say
viii
without hesitation that I would take this path again even if I were given a second
chance is because I was surrounded by wonderful family and friends who made
what could have been the most stressful time of my life into unforgettable five years
filled with cherished moments, friendship, love, and laughter.
First, a big thank you to all of my friends in Korea. Although they could not
be here with me in body, I could not have pulled through without their constant
support via IMs, various social networking sites, long international telephone
conversations, and good ol’ hand-written letters. Thank you for being there for me
through thick and thin. I would also like to thank all of my Korean friends here at
Stanford who kept me well fed and pampered. I feel that I have received so much
more than I have given during the past five years and will never forget the
incredible social support that all of you have shown me. It was amazing how every
time that I felt I could not go on and wanted to pack my bags, there was someone
out there ready to hold my hand and tell me that this, too, would all pass. I will
never forget the experiences I had as the President of the Korean Student
Association at Stanford, the late night movies and drinks, the potluck parties, and
the countless hours spent pouring our hearts out over coffee and tea.
An extra special thanks to the members of JAGH – Jesse Fox, Abhay
Sukumaran, and Helen Harris – for their incredible support and friendship. You
guys are the best cohort that a grad student could wish for. This dissertation could
never have happened without your words of advice and encouragement, warm
brownies, surprise gift and flower deliveries. I expect to see all of you in Georgia
very soon. Thank you, Kathryn Segovia, for having more faith in me than I have in
myself. You always tell me that I am a Superwoman, but I could never have done it
without your happy smiles, cheesecake deliveries, amazing baby showers, and
ix
constant encouragement. You brought out the Superwoman in me. I would also like
to thank Sean Westwood for his wry sense of humor and awesome cooking skills,
and Ethan Plaut for sharing my quirky likes and dislikes of life.
A giant hug and kisses to my parents, Kyu-Moon Ahn and Youn-Sook Cho,
for being the best parents anyone could possibly want. I took it for granted for quite
a while but now realize how incredible it is that there is not a thing I would change
about my childhood or the way that I was brought up. You have always been the
source of my strength and I just know that you will make everything better even in
the worst of times. The greatest obstacle to obtaining my doctoral degree was the
fact that I was only able to see you once or twice a year at most. And a great big
bear hug to my little brother, Sun-chong (Richard) Ahn. Thank you for being so
dependable even though you are my dongseng.
Finally, I would like to dedicate this dissertation to my husband, Yeon Jin
(Andrew) Chu, and my soon-to-be-born son, Ryan Chu. I love both of you more
than anything else in the world. Yeon Jin oppa, you always tell me that the best
decision you ever made in your life was to marry me. I have just three words in
reply to that: back at you.
x
TABLE OF CONTENTS
CHAPTER ONE: INTRODUCTION………………………………………………1
CHAPTER TWO: EMBODIED COGNITION AND EMBODIED
EXPERIENCES……………………………….…………………………………….4
Neurological Evidence of Embodied Cognition…………………………..........4
Behavioral Evidence of Embodied Cognition………………………………….6
Mental Simulation………………………………………………………………8
Virtual Environments and Embodied Experiences……………………………..9
CHAPTER THREE: FUNCTIONAL FEATURES OF EMBODIED
EXPERIENCES IN VIRTUAL ENVIRONMENTS………………………………12
Immersion…………………………………………………………………..…12
Interactivity………………………………………………………………...….15
Combination of Flexibility, Control, and Realism…………………………….17
CHAPTER FOUR: SUBJECTIVE FEATURES OF EMBODIED EXPERIENCES
IN VIRTUAL ENVIRONMENTS………………...……………………………….22
Individual Differences in Presence Perception………………………………..27
Presence Perception in Different Modalities………………………………….29
CHAPTER FIVE: JUSTIFICATION FOR DISSERTATION STUDIES………….33
Different Modalities and Embodied Experiences………………………….….33
Processes of Embodied Experiences…………………………………………..37
Individual Differences in Embodied Experiences…………………………….38
CHAPTER SIX: DISSERTATION STUDY ONE: COMPARISION OF
EMBODIED EXPERIENCES WITHIN IVE AND MENTAL SIMULATION.......41
Methods…..……...………………………………………………………..44
Results…......……………………………………………………………....54
xi
Discussion……………………………………………………………………..58
CHAPTER SEVEN: DISSERTATION STUDY TWO: EFFECT OF IMMERSION
AND AGENCY OF CONTROL ON EMBODIED EXPERIENCES……………..61
Methods………………………………………………………………………..64
Results………………………………………………………………………....71
Discussion……………………………………………………………………..76
CHAPTER EIGHT: CONCLUSION………………………………………………79
Summary of Findings………………………………………………………….79
Theoretical Implications – Embodied Cognition……………………………...80
Theoretical Implications – Immersion and Presence in Virtual Environments
…………………………………………………………………………………84
Theoretical Implications – Environmental Studies……………………………86
Alternative Theoretical Explanations...…………………………….………….87
Applied Implications………………………………………………………….90
Limitations and Future Directions…………………………………………….92
NOTES……………………………………………………………………………..97
APPENDIX A: STIMULUS FOR MENTAL SIMULATION CONDITION IN
STUDY 1…………………………………………………………………………..98
APPENDIX B: STUDY 1 SURVEY ITEMS...…………………………………..100
APPENDIX C: SURVEY ITEMS FOR STUDY 2……………………………….104
APPENDIX D: STATISTICS FOR ALL ANALYSES IN STUDY 1…………...108
APPENDIX E: MEANS AND STANDARD DEVIATIONS FOR ALL
DEPENDENT VARIABLES IN STUDY 2………………………………………110
APPENDIX F: STATISTICS FOR ALL ANALYSES IN STUDY 2……...……120
REFERENCES……………………………………………………………………124
xii
LIST OF TABLES
Table 1. Study 1 Means and Standard Deviations for Dependent Variables by
Condition…………………………………………………………………………...55
Table 2. Pearson Correlations between Dependent Variables in Study 1 (N = 47)
……………………………………………………………………………………...57
Table 3. 2x2 Between-Participants Design for Study 2……………………………66
Table 4. Pearson Correlations between Dependent Variables in Study 2 (N = 101)
……………………………………………………………………………………...75
xiii
LIST OF FIGURES
Figure 1. Experimental Setup for the IVE Condition in Study 1…………………..48
Figure 2. Screenshot of the Series of Events in the IVE Condition in Study 1……51
Figure 3. Experimental Setup for Low Immersion, Self-Move Condition in Study 2
……………………………………………………………….……………………..67
Figure 4. Interaction between Agency of Control and Perspective Taking Propensity
on Napkin Use in Study 2………………………………………………………….72
Figure 5. Interaction between Self-Efficacy and Perspective Taking Propensity Over
Time in Study 2…………………………………………………………………….74
xiv
CHAPTER ONE: INTRODUCTION
Just as the Internet has significantly changed the way that we express
ourselves and interact with others (Bargh & McKenna, 2004), the advancement of
digital media such as video games and virtual environments continues to transform
traditional communication rules. Researchers from various fields have discovered
numerous benefits of applying state-of-the-art digital media to their work as
research tools (Blascovich et al., 2002; Persky & McBride, 2009). Using digital
technology, scholars are able to obtain reliable measurements of human behavior
that are accurate to the millisecond and millimeter while maintaining a balance
between ecological validity and experimental control which has long plagued
empirical scientists as a shortcoming of laboratory research.
In addition, novel affordances of digital media give rise to unique
approaches and solutions to research themes of the past that could not be answered
with traditional methods, due to the lack of technical capacity needed to approach
such an issue. This dissertation explores how advanced digital technology can
augment human sensory capacities (e.g., Bailenson, Beall, Blascovich, Loomis, &
Turk, 2005) and how digitally enhanced multisensory experiences can influence
attitudes and behaviors that transfer to the non-mediated, physical world.
Furthermore, unique affordances of these advanced digital media are parsed out and
investigated to gain deeper understanding of the underlying mechanisms of digitally
augmented experiences.
The capacity to digitally augment sensory experiences allows advanced
digital technology to offer embodied experiences – a vivid, realistic mediated
experience bolstered by simulated sensory information. Mediated experiences were
available with more traditional media such as books and television (Reeves & Nass,
1
1996) but novel affordances of advanced digital media allow users to go beyond
passive viewership and become active participants in the mediated context, seeing,
hearing, and feeling the experience as if it were their own. The questions of interest
in this dissertation are how active participation in the mediated experience
compares to the more passive interactions with traditional media and the specific
elements of advanced digital media that influence user attitudes and behaviors.
Immersion is the main novel affordance that enables digital media to
reproduce realistic and vivid embodied experiences. Conceptually, immersion refers
to users being surrounded or enveloped by sensory information simulated by digital
devices that allow them to temporarily forget the fact that they are in a mediated
environment, causing them to think and behave as if they are in the real, physical
world (Heeter, 1992; Slater, Usoh, & Steed, 1994; Steuer, 1995). Immersion is one
of the affordances that differentiate advanced digital media from traditional ones in
that they are able to mimic the real world and face-to-face interactions with more
realism than any other traditional medium to date (Grigorovici, 2003).
The realism of embodied experiences is important because the theory of
embodied cognition stipulates that our mind is defined by the limitations of our
body and our specific situation or environment (Clark, 2001; Gallagher, 2005). If
we learn and think via the experiences involving our body, then the simulated
experience becomes similar to experiences in the non-mediated, physical world.
The advantage of using advanced digital media to explore embodied experiences is
that there is almost infinite flexibility in terms of the mediated context constructed
for these simulated experiences. Thus, users are able to actively participate in
experiences that they were only able to imagine with traditional media, while
receiving vivid, tangible sensory inputs. This makes it possible for scholars to
2
answer innovative and perhaps radical questions with regard to how active
experiences can influence the way individuals think and behave.
Despite the promising potential for real life implications and benefits,
surprisingly few empirical studies have looked at applying advanced digital media
such as virtual environments to explore the processes and effects of embodied
experiences. This dissertation aims to lay the cornerstones to study the use of virtual
environments in embodied cognition research and the significance of immersion as
a novel affordance of advanced digital media. The following chapters will 1)
provide an overview of embodied cognition theory and introduce the concept of
embodied experiences; 2) discuss functional characteristics of embodied
experiences through digital media; 3) discuss subjective characteristics of embodied
experiences through digital media; 4) present two new studies designed to compare
embodied experiences against imagined experiences and how these modalities
influence attitude and actual behaviors in the physical world; and 5) discuss
theoretical and applied implications of embodied experiences using advanced
digital media.
3
CHAPTER TWO: EMBODIED COGNITION AND EMBODIED EXPERIENCES
The theory of embodied cognition can be distinguished from more
traditional cognitive science in the emphasis it places on the role of the body – body
being the biological component of self (versus the mind) – and the surrounding
environment. Earlier cognitive theories thought of the cognitive process as
analogous to a computer’s main processing unit, assuming that the mind is software
that runs on brain hardware and cognition is a logical process based on pure
computation. In comparison, the theory of embodied cognition focuses on the
physical attributes of the body and interactions among the body, environment, and
the mind (Shapiro, 2007; Wilson, 2002). For instance, when infants learn how to
perform a simple action such as picking up an object, they do so through dynamic
interactions with their environments and discover the best way to execute this action
through trial and error rather than follow some innate, pre-programmed neural map
inside their brains (Thelen & Smith, 1994).
Neurological Evidence of Embodied Cognition
Embodied cognition emphasizes intrinsic ties between the body’s action
and cognitive processes (Feldman & Narayaman, 2004) and opposes the view that
the mind moves the body through abstract and formal logic. The discovery of mirror
neurons provides a biological explanation for this action-perception link. Mirror
neuron activation was discovered in the premotor cortex of monkeys, exhibiting
increased activity both when a monkey performed hand movements itself and when
it merely observed another monkey or a human performing the same hand
movements (Ferrari, Gallese, Rizzolatti, & Fogassi, 2003; Gallese, Fadiga, Fogassi,
& Rizzolatti, 1996). Similar neuron activity was found in the human brain using
functional neuroimaging studies (Grezes & Decety, 2001), implying that perception
4
is linked with action inside the brain. That is, once sensorimotor schemata that link
the body and mind are constructed in our minds, merely perceiving or imagining
another person in a situation activates neurons connected with the actual action. For
instance, the same part of the brain became activated both by merely viewing
pictures of disgusted faces and by actually smelling disgusting odors (Wicker et al.,
2003). Mirror neuron activity was also triggered when only the sounds that a certain
action produces were heard (Kohler et al., 2002).
Further evidence has been found by Rizzolatti and colleagues (Di
Pellegrino, Fadiga, Fogassi, Gallese, & Rizzolatti, 1992) with the discovery of
canonical neurons in the rear section of the arcuate sulcus. Canonical neurons
respond selectively to three-dimensional objects, differentiating them in function of
their shape, size, and spatial orientation. For example, if an object is large enough
for an individual to use his or her whole hand to grasp it (versus using just fingers),
mere observation of an object of similar size will activate the same canonical
neurons. However, these neurons will not activate upon observation of a smaller,
different sized object. It should be noted that these neurons will be activated not
only in response to observing same object that the individual has interacted with,
but with objects that require similar interactions.
The activation patterns of these neurons indicate that once sensorimotor
schemata are constructed, mere perception of related objects or movements activate
relevant neuron systems. Thus, once the body learns how to interact in a certain way
with the environment, observation of objects with similar characteristics
automatically evokes the most suitable motor program required to interact with the
environment (Gallese, 2000). This process of simulation is what mediates the
relationship between action and perception (Lakoff & Johnson, 1999), and
5
neurological evidence confirms that simulation is not based on logic and reason but
on sensorimotor experiences.
Behavior Evidence of Embodied Cognition
Behavioral evidence of the body influencing how individuals think and feel
can also be found. Objects and events that produce similar affective responses are
remembered together in the same perceptual category (Niedenthal, Halberstadt, &
Innes-Ker, 1999), and these perceptual inputs are later used as bases in future
judgments (Barsalou, 1999). More recent studies showed that individuals primed
with either hot or cold tactile temperatures by holding hot or cold drinks before
meeting another person would later judge that person to have a warm or cold
personality, respectively. The same manipulation also led individuals to behave in a
prosocial or egocentric manner, respectively (Williams & Bargh, 2008). Participants
in both studies demonstrated no awareness of the impact of the physical experience
on their judgment. Thus, tactile experiences of physical warmth or coldness are able
to directly influence the way people think, feel, and behave even when they are
unconscious of this influence. Similarly, the sensory experience of feeling the
relative weight of objects can affect judgments. When holding a heavy clipboard,
participants assessed foreign currency as more valuable compared to those holding
a lighter clipboard (Jostmann, Lakens, & Schubert, 2009). Also, the perception of
darkness created by wearing sunglasses (as opposed to clear glasses), has been
shown to increase the perception of anonymity, leading to increases in morally
questionable actions such as cheating (Zhong, Bhons, & Gino, 2010).
If the above studies demonstrated the link between sensory perception and
cognition, other studies evidenced the connection between motor perception and
cognition. People are likely to agree with statements when they make nodding head
6
movements while listening (Wells & Petty, 1980); rate a cartoon as humorous when
their facial muscles used for smiling are stimulated unobtrusively as they view the
cartoon (Strack, Martin, & Stepper, 1988); feel pride in an achievement when sitting
upright than slumped as they hear news about their performance (Stepper & Strack,
1993); and find a random pair of letters likeable if they have extensive experience
typing and the letter pair is easy to type (Beilock & Holt, 2007).
Body movements have also been shown to influence higher order cognition,
such as solving complex physics problems. When participants were asked to solve a
physics problem and either asked to make arm movements relevant to the solution
or irrelevant to the solution, those who made relevant arm movements were
significantly more successful at solving the problem even when they were unaware
that the arm movements were relevant to the solution (Thomas & Lleras, 2009).
Thus, leading individuals to move in a particular way can influence the processes of
the mind. On the other hand, limiting motor perceptions can also restrict the
processes of the mind. In a recent study, when facial muscles used in frowning were
paralyzed with Botox injections, participants took a significantly longer time to read
sentences evoking anger that would trigger frowning muscles compared to reading
sentences that evoke sadness or happiness (Havas, Glenberg, Gutowski, Lucarelli,
& Davidson, 2010).
It should be noted that the action-perception link demonstrated in these
studies are different from behavioral modeling such as that of social cognitive
theory (Bandura, 1977), which stipulates that people can learn behaviors vicariously
through observation of models. Rather, the sensorimotor experience – nodding,
holding a warm drink – seem to trigger the sensorimotor schema related to the
experience in some way – perception of agreement, perception of kindness. Due to
7
the relatively short history of research on embodied cognition, the exact underlying
mechanism that links the sensorimotor experience to relevant sensorimotor schema
is not entirely clear. However, neurological and behavioral evidence discussed
above demonstrate that humans have the ability to categorize mental schema into
relevant groups, and a sensorimotor experience can trigger any relevant schema
categorized into the same group. This is why individuals relate the experience of
visual darkness to anonymity and immoral actions, and how they link physical
warmness to warm personalities although the sensorimotor experience does not
necessarily connect directly with the triggered schema.
Mental Simulation
As discussed above, once sensorimotor schemata are created, even the mere
observation of related objects, events, and people activate neurons involved with the
action. Upon observation, relevant neural systems are activated as if the observer
were interacting with it. This virtual, or “off-line,” simulation of potential action is
known as mental simulation (Goldman, 1989). Because of this link between mental
simulation and action neurons, mental simulation techniques are commonly used to
develop athletic skills (Hausenblas, Hall, Rodgers, & Munroe, 1999) as well as
promote task self-efficacy and enhance skill performance (Cumming, 2008; Martin
& Hall, 1995).
Several studies have explored the effect of mental simulation on attitude and
behavior change to demonstrate that simple imagination is able to trigger perceptual
simulation of the senses based on previously constructed sensorimotor schema. For
instance, participants who recalled a past experience of social rejection perceived
the room’s temperature as significantly lower than participants who recalled
experiences of social inclusion (Zhong & Leonardelli, 2008). Similarly,
8
retrospective thinking about life in the past triggered participants to lean backwards
whereas prospective thinking about life in the future triggered them to lean forward
(Miles, Nind, & Macrae, 2010).
A series of other studies show that text stimuli are able to induce mental
simulation that leads to activation of sensorimotor schema. When reading verbs
related to either positive or negative emotions (e.g., to smile, to frown), facial
muscles associated with positive or negative expressions were stimulated (Foroni &
Semin, 2009). The connection between text-induced mental simulation and the
activation of sensorimotor schema is further supported by neurological evidence
from right- and left-handed individuals. When reading verbs related to manual
action (e.g., grasp or throw), the area of the brain that is triggered during actual
motor activity for right- and left-handed individuals became active, respectively
(Willems, Hagoort, & Casasanto, 2009). A related neurological study demonstrated
that individuals understand a story by mentally simulating different events in the
narrative (Speer, Reynolds, Swallow, & Zacks, 2009).
Virtual Environments and Embodied Experiences
The fact that people’s thoughts and feelings are limited by their body’s
physical capacities seems to be a big constraint of embodied cognition (Glenberg,
1999). That is, human action and thoughts are bound by the affordances of the
human body, such as body type, perceptual apparatus, and means of interaction with
the environment. For instance, a human being has completely different means of
playing with a ball compared to a puppy and thus constructs very different
sensorimotor schema that link actions with cognitions. Mental simulation would
also be bound by this restriction even if it is based on imagination – without a
previously constructed sensorimotor schema, there would be no basis to guide the
9
simulation. This implies that simulating a situation that is beyond the capacities of
the human body or that the individual has no prior experiences may be difficult to
process.
Virtual environments offer unique solutions so that individuals may go
beyond the constraints of the body or of prior experiences. At the click of a button,
users are able to change their self-representation at will to any type or form of body,
sex, or species and experience the virtual world embodying a virtual representation
completely different from the physical one in the real world. Furthermore, virtual
environments are able to augment the sensory capacities of the human body, for
instance, by allowing users to simultaneously expand their gaze to different people
or to be present at two different places at the same time (Bailenson, Yee, Blascovich,
& Guadagno, 2008).
Although more traditional media such as television, film, and video have
been able to extend the capacities of human sensory systems to some degree (e.g.,
show the world from another person’s point of view, present fantastical characters
that viewers can mentally embody), virtual environments offer novel affordances
that maximize the experience of embodiment. In this dissertation, embodied
experiences are defined as the users’ experiences within virtual environments with
vivid visual, aural, and haptic sensorimotor inputs and the ability to actively interact
with the mediated environment, such as with objects and people. These are
experiences in which users are able to embody a completely different body and see,
hear, and feel completely novel situations that transcend physical space and time.
Such experiences have not been possible with traditional media and present
immense potential to gain deeper insight into the processes of the body and mind by
asking questions that have been difficult to address until recently. Embodied
10
experiences in virtual environments are as close as one can get to physical
experiences using mediated content (Grigorovici, 2003), but they are still virtually
mediated experiences. Thus, the question of interest is whether sensorimotor
experiences in virtual worlds are able to influence attitudes and behaviors in the
physical world. Also of interest is how embodied experiences compare to mental
simulation of the same experience when individuals lack a prior constructed
sensorimotor schema.
To better understand the characteristics and affordances of virtual
environments and embodied experiences, a discussion of functional and subjective
features of virtual environments that make embodied experiences possible is in
order.
11
CHAPTER THREE: FUNCTIONAL FEATURES OF EMBODIED
EXPERIENCES IN VIRTUAL ENVIRONMENTS
Virtual environments are realistic artifacts simulated by digital computers,
devices, and their associated technologies that blur the distinction between reality
and its virtual representation (Ellis, 1995). There are several functional features of
virtual environments that set them apart from traditional media in terms of
providing embodied experiences. These affordances augment sensorimotor inputs
and aid in creating experiences that are realistic enough for the user to respond as if
he or she is in a non-mediated, physical environment.
Immersion
Immersion is a concept often used to describe a novel affordance of digital
media which serves to heighten the reality of the mediated content. Slater and
Wilbur (1997) define immersion as the extent to which media are capable of
delivering an inclusive, extensive, surrounding, and vivid illusion of reality to the
senses. Singer and Witmer (1999) describe immersion as an individual’s perception
and reaction to a virtual environment. Murray (1997) suggests a more metaphorical
definition of immersion: “Immersion is a metaphorical term derived from the
physical experience of being submerged in water. We seek the same feeling from a
psychologically immersive experience that we do from a plunge in the ocean or a
swimming pool; the sensation of being surrounded by a completely other reality…
that takes over all of our attention, our whole perceptual apparatus…” (p. 98-99).
Although the field has not been able to settle on a canonical definition yet, taken
together, the concept of immersion describes how users are surrounded by layers of
sensory information simulated by advanced digital technology that create the
illusion of being in the physical world.
12
Virtual environments can differ in the level of immersion that they offer.
Immersive virtual environments (IVEs) are virtual environment systems that
amplify the effect of simulation by surrounding the user with numerous layers of
sensory and perceptual information created by digital devices (Loomis, Blascovich,
& Beall, 1999). The more sensory information is provided, the higher the level of
immersion that the IVE offers. IVEs are able to surround users with stereoscopic
visual input, spatialized aural input, and tactile sensory input, yielding a sense of
perceptual depth and vivid realism to the virtual world. For instance, by building a
simulator which represents all the sensory perceptions experienced by a
schizophrenic, an ordinary person is able embody the realistic perceptual
experiences of this mental disability (Kalyanaraman, Penn, Ivory, & Judge, 2010).
In comparison, virtual environments with lower levels of immersion such as
desktop computers offer monoscopic visual, simple audio, and limited tactile
sensory input.
IVEs can be implemented in different ways (see Loomis, Blascovich, &
Beall, 1999 for details). The IVE system used for this dissertation was comprised of
the head-mounted display (HMD), a headpiece with a lens for each eye which
provides stereoscopic views of the computer-generated environment, and various
devices that track simple head and body movements as well as the position of the
body in three-dimensional space. Spatialized sound from headphones built into the
HMD allowed users to hear sounds that seemed to be emanating from surrounding
auditory space (Wenzel, 1992). A force feedback haptic device was applied to yield
tactile stimulation of arm and hand movement and interaction with the virtual
environment.
Several studies have explored how the scope and range of sensorimotor
13
inputs provided by the virtual environment system can influence users’ attitudes and
behaviors. One of the earliest efforts to quantify the levels of immersion of different
devices compared head-controlled point of view against hand controlled point of
view in a search task and found that users with head-controlled point of view spent
substantially less time compared to those with hand controlled point of view
(Pausch, Proffitt, & Williams, 1997). Other studies have also compared higher (i.e.,
IVEs) and lower (i.e., desktop computers) immersive virtual environment systems
and confirmed significant benefits of IVEs in terms of data visualization and
processing (Arns, Cook, & Cruz-Neira, 1999; Raja, Bowman, Lucas, & North,
2005); performance on a path construction task (Gruchalla, 2004); search and
navigation task (Boyd, 1997); depth and distance perception (Heineken & Schulte,
2000); collaboration task involving close interaction with a remote user (Roberts,
Wolff, & Otto, 2005); configuration and movement of chess pieces (Slater, Lanikis,
Usoh, & Kooper, 1996); and short-term spatial memory when a navigation task is
involved (Johnson & Adamo-Villani, 2010). For example, various levels of
immersion in a virtual environment were found to be more effective in terms of
directional accuracy and precision locating an unfamiliar landmark compared to
static images on paper or paper and pencil drawings (Waller, Beall, & Loomis,
2004).
Some studies report no significant differences between high and low
immersive environments. In a virtual search and navigation task presented in either
HMD or desktop, Slater and colleagues (1995) failed to find significant differences
between the two presentation modes. A pilot study (Wiederhold, Davis, &
Wiederhold, 1998) observed responses from five participants after presenting a
virtual airplane ride through a HMD or a 3D-monitor but found no differences in
14
physiological responses such as heart rate, skin temperature, and respiration rate
between the two presentation modes.
Despite some inconsistent findings, these studies generally imply that
higher IVEs which provide more numerous layers of perceptual inputs may be more
effective than lower IVEs, particularly for sensorimotor tasks. For the dissertation
studies, it is anticipated that virtual environments which provide higher levels of
immersion through rich layers of perceptual input will construct more influential
embodied experiences compared to lower immersive stimuli. As a result, embodied
experiences offered via high IVEs will be more likely than lower IVEs to alter
attitudes and behaviors in the non-mediated physical world.
Interactivity
Interactivity is another characteristic of virtual environments that provide
sensorimotor experiences, and is defined as the process in which the user can
influence the form and/or content of the mediated experience (Heeter, 2000;
Lombard & Ditton, 1997); the user’s perception that the medium is responding to
the user’s actions (Leiner & Quiring, 2008; Steuer, 1995); and the medium’s
technological capacity to actually respond (Lee, Park, & Jin, 2006) in a two-way
exchange, immediately, in real-time (Rice & Williams, 1984). It is generally agreed
that the best interactive medium mimics the interactive dynamics of face-to-face
communication (Rafaeli, 1988). While interactivity is not a characteristic unique to
virtual environments, virtual environments tend to provide users with the most
engaging and responsive stimuli that respond in real-time to user actions.
Interactivity allows the user to be both an observer and a participant in
mediated environments, possibly leading to more potent media effects (Vorderer,
2000). Virtual environments offer interactivity in several ways that encourage
15
greater user control, participation, and engagement (Sundar, Xu, & Bellur, 2010).
On websites, input and output features built into the medium such as scroll bars,
hyperlinks, or movie clips keep many sensory channels involved during an
interaction between media and their users. Tailored content that offer enhanced
personalization over time based on user inputs (e.g., news feeds, avatar creation in
video games) allows users to become not just passive receivers, but the source and
the sender during the communication process. In IVEs, interactivity allows users to
move in the virtual environment as if they were in a non-mediated, physical world
and the objects and people within the IVE aim to respond to users as they would in
the real world.
There are various benefits of higher interactivity (compared to lower
interactivity) such as better product knowledge and positive brand attitudes after
exposure to online advertisements (Ahn & Bailenson, in press; Li, Daugherty, &
Biocca, 2002), higher audience involvement compared to traditional media (Fortin
& Dholakia, 2005), greater perception of reality (Biocca et al., 2001; Edwards &
Gangadharbatla, 2001; Li, Daugherty, & Biocca, 2001; Lombard & Ditton, 1997;
Skalski & Tamborini, 2007), and more positive attitude or liking toward the
mediated content (Sundar & Kim, 2005).
On the other hand, multi-sensory stimuli that offer immediate feedback and
response to user inputs may incur unforeseen costs such as cognitive overload.
Presenting individuals with highly interactive advertisements with animation
features degraded product memory although it increased recognition of the
advertisement (Sundar & Kim, 2005). Other studies demonstrate that although users
like audio and video features of a website, they learn more through the least
interactive modality in the form of text and static pictures (Sundar, 2000). Contrary
16
to what intuition might suggest, more interactive features is not necessarily better.
Although these features might capture the users’ attention or interest, it is also
highly likely that numerous interactive features will tax their cognitive resources.
Thus, a timely investigation of the effect of interactivity presented in virtual
environments – especially when coupled with immersion – on user perceptions of
embodied experiences as well as its effects on people’s attitude and behavior in real
life is imperative. Interactivity features presented in IVEs are based on naturalistic
tracking of movements and gestures and may be more intuitive compared to virtual
environments that are lower in immersion, such as websites. That is, interactive
features within IVEs such as head-controlled point of view and haptic feedback are
designed to mimic real perceptions from the physical world. Thus, much of the
interactive features presented in high IVEs rely on existing sensorimotor schemata.
To this extent, it is anticipated that even with highly interactive features, embodied
experiences within IVEs will not place a burden on the users’ cognitive load.
Combination of Flexibility, Control, and Realism
Another important functional feature of virtual environments that would be
appealing for scholars studying embodied experiences is that the flexibility of the
context of the experience is almost limitless. Within virtual environments, users
have the freedom to embody any type of body and be inside any type of
environment. The flexibility of body type is made possible through the use of
avatars, a form of virtual self-representation that the user is able to control within
the virtual environment (Ahn, Fox, & Bailenson, in press). The flexibility of
environment type is made possible through the flexible nature of computer graphics
that allow users to render realistic representations of the physical world. The
distinctive advantage of virtual environments is that in addition to this flexibility,
17
individuals have vivid and tangible sensorimotor experiences with high immersion
and interactivity as if they were in a non-mediated, physical world.
A clear example of this balance between flexibility and high realism is
Lanier’s concept of homuncular flexibility (2006). The homunculus is the mapping
of the body into the motor cortex. The notion of its flexibility posits that humans are
able to embody virtual self representations that have completely different body parts
with limbs they have never used before. Lanier’s example of homuncular flexibility
draws upon embodying a virtual lobster with more than two arms and two legs.
Despite the lack of previously constructed sensorimotor schemata on how to move
the extra six limbs, some practice within interactive IVEs can help humans learn to
adapt to their virtual body parts. The concept of this “homuncular lobster” implies
that embodied experiences within IVE are able to expand the human mind to
construct completely novel sensorimotor schemata and connect those schemata to
completely novel body parts. In essence, IVEs allow users to experience a literal
translation of “stepping into the skin” of another being, and high immersion and
high interactivity make this embodied experience sufficiently natural and realistic
for individuals to respond as if the embodied experience were occurring in the real
world.
A series of experiments demonstrated IVEs’ capacity to allow individuals to
virtually step into another person’s body and experience the world through that
person’s eyes to explore the effect of embodiment on their attitude and behavior. By
embodying the body of an elderly avatar in an IVE, participants demonstrated a
decrease in negative stereotypes against elderly individuals (Yee & Bailenson,
2006). When participants embodied an attractive avatar, they were more willing to
stand closer to an unfamiliar person and disclose more information about the self
18
compared to those who embodied less attractive avatars (Yee & Bailenson, 2007).
In the same paper, participants who were assigned taller avatars negotiated more
aggressively than those who were given shorter avatars on a negotiation task. In a
follow-up study, Yee and colleagues (2009) demonstrated that the effect of
embodying tall or short avatars in the virtual world could transfer into the physical
world, where participants who embodied taller avatars negotiated more aggressively
than those who embodied shorter avatars. Groom and colleagues (2009) had
participants embody either an African-American or a White Caucasian avatar in the
virtual world and asked them to “imagine a day in the life of this individual as if
you were that person.” They demonstrated that the experience of virtually
embodying another person’s body was able to influence people’s attitude about
racial bias (through stereotype activation) whereas mentally simulating the
embodiment failed to do so.
IVE also allows users extensive control of the mediated environment in that
they can easily recreate any type of environment as needed, transcending spatial
restrictions. At the touch of a button, users can stand in a classroom or be on an
island, with vivid visual, aural, and tactile sensory inputs from each environment as
well as the ability to interact with the objects in the environment. Also, just like
other advanced digital media such as video, IVEs are also able to accurately
replicate and reproduce the same stimuli almost infinitely and users can experience
the same experience time and time again.
Experimenters can benefit from IVE’s capacity for extensive control in that
they are able to unobtrusively track and measure every movement that the user
makes in the virtual environment. Thus, they have begun to view IVEs as optimal
solutions for exercising experimental control and obtaining accurate measurements
19
while retaining mundane realism in the laboratory (Blascovich et al., 2002).
Numerous researchers have used IVEs to explore and replicate traditional social
psychological theories. Among many others, interpersonal distance (Bailenson,
Blascovich, Beall, & Loomis, 2003), social facilitation (Hoyt, Blascovich, & Swinth,
2003), and prejudice (Dotsch & Wigboldus, 2008) theories studied in the physical
world have been replicated successfully within IVEs. In health contexts, Persky and
McBride (2009) noted the significant potential of IVEs to be integrated into medical
training and services by having clinicians embody the viewpoint of their patients to
experience and evaluate their own counseling technique. Furthermore, disabilities
such as schizophrenia (Kalyanaraman, Penn, Ivory, & Judge, 2010) and various
visual disabilities (Jin, Ai, & Rasmussen, 2005) have been accurately reproduced as
virtual simulations. Although the technology is far from perfect and even IVEs with
the highest degree of immersion are still unable to completely mimic the real world,
these studies confirm that IVE technology can invoke naturalistic responses from
participants while maintaining extensive experimental control.
In sum, IVEs provide a novel approach for the embodied cognition research
program by allowing individuals to virtually embody an avatar of any type or form
in any sort of environment. During the experience, users are surrounded by rich
layers of sensory inputs to see, hear, and touch objects in a three-dimensional space
that responds in real time to user inputs. Compared to traditional media or mental
simulation, this is a much more tangible and interactive form of embodied cognition
particularly when the individual has no prior constructed sensorimotor schema on
the particular body type (i.e., embodying an elderly person) or the environment (i.e.,
living in a foreign country). Details of the embodied experience may be tailored to
the individual, for instance, sensory inputs that are programmed to the millimeter
20
and millisecond to control exactly what the user sees, hears, and feels. Thus, IVEs
offer an optimal combination of flexibility, control, and realism, potentially making
them attractive tools to explore questions of embodied cognition.
21
CHAPTER FOUR: SUBJECTIVE FEATURES OF EMBODIED EXPERIENCES
IN VIRTUAL ENVIRONMENTS
Given that the ultimate goal of all media is to reproduce the illusion of nonmediated communication as best as they can, the degree of vividness or realism that
users experience during their experiences within virtual environments will influence
their attitudes and behaviors. Presence is the psychological state of feeling that the
mediated experience reproduced by virtual environments is “real.” The term
telepresence, which preceded the widespread use of “presence” as a more
generalized application of the concept, was first coined by Marvin Minsky (1980) to
refer to the sense of being physically present at a remote location through
interaction with a system’s interface. Biocca (1997) went on to refine the definition
of presence, referring to it as the illusion of “being there” regardless of whether
“there” is physical, mediated, or imagined. Lombard and Ditton (1997) presented a
similar definition of presence as the perceptual illusion of nonmediation, meaning
that although individuals can later indicate that they were using the technology,
their subjective perceptions make it seem as if the technology was not involved in
the experience (Lombard, 2001).
Some scholars have used immersion and presence interchangeably because
of their obvious conceptual similarities. In this dissertation, the two concepts will be
distinguished by defining presence as the subjective unit of evaluating the realism
of the mediated experience, whereas immersion would be the objective, or
technological unit of evaluation. That is, even when users are exposed to the same
mediated environment (i.e., immersion), individuals may have different subjective
assessments of how realistic it is (i.e., presence). Immersion will be operationalized
by manipulating the components and layers of sensory inputs provided by digital
22
technology while presence will be measured as various forms of individual
assessments of the virtual environment’s realism.
Because presence is an entirely subjective experience of reality while
exposed to mediated content, users’ perception of presence is bound to change over
time and the advent of sophisticated technology. For example, when the Lumiere
brothers introduced the first motion picture in 1895 of a train approaching a station,
it had people screaming and running for cover. Obviously, the audience back then
felt a high sense of presence while viewing the movie. However, two-dimensional
black and white films are no longer able to elicit such high levels of presence
among today’s audience as their reception and understanding of mediated content
have become more sophisticated over time (Campbell, 2000). As a result, it is
expected that relatively newer forms of media which offer richer sensory inputs will
be more successful in terms of creating realistic illusions of “being there.” Green
and colleagues (2004) also note that newer forms of digital media such as video
games or virtual reality simulation are particularly more likely to result in high
presence compared to traditional media.
The vivid sense of “being there” created through perception of presence
offers users with experiences realistic enough to lead to persuasion and attitudinal
changes. For instance, consumers with high perception of presence were more
likely to be persuaded by a television advertisement compared to those who felt less
presence (Kim & Biocca, 1997). In their study, this difference was driven by the
subjective experience of presence alone and the manipulation of immersion (i.e.,
angle of viewing, lighting) did not affect results. A study exploring video game
research found that IVE and desktop platforms induced different levels of presence,
and the perception of presence mediated the likelihood to develop aggressive
23
feelings after playing a violent video game – high presence led to more aggressive
feelings (Persky & Blascovich, 2008). In the realm of health research, high levels of
perceived presence induced by an interactive computer agent led individuals to
react favorably to the agent when the agent was attractive and even reduced
negativity when the agent was unattractive. Furthermore, when perceived presence
was high (i.e., higher interactivity), individuals were more likely to be persuaded by
health messages compared to when perceived presence was low (Skalski &
Tamborini, 2007). Similarly, high perception of presence elicited by interactive selfrepresentations led to changes in eating behaviors (Fox, Bailenson, & Binney, 2009).
Studies investigating task performances have found that high levels of perceived
presence are also positively linked with task performance (Sas, O’Hare, & Reilly,
2004; Slater, Linakis, Usoh, & Kooper, 1996).
These studies imply that the subjective perception of presence may be the
underlying mechanism driving the success of navigation and task performance
within virtual worlds. Another implication is that greater perception of presence
during the embodied experience may lead to important influences on users’ attitude
and behavior in the non-mediated, physical world. Intuitively, because embodied
cognition theory stipulates the connection between sensorimotor experiences of the
body and the mind, it is likely that this relationship would be stronger if the user
perceives the sensorimotor experience in the virtual world to be more realistic.
Indeed, studies have demonstrated that when the sensory experience is sufficiently
realistic, the person feels an illusion of the physical body merging with the body of
the self representation in the virtual environment (Sanches-Vives, Spanlang, Frisoli,
Bergamasco, & Slater, 2010; Slater, Spanlang, Sanches-Vives, & Blanke, 2010). In
one of these studies, Sanches-Vives and colleagues demonstrated that by
24
synchronizing the haptic sensation of moving the physical hand and the hand of the
self avatar for approximately 2-3 minutes, participants later felt the sensation of
their stationary physical arm moving when they saw the virtual arm move. This
sense of embodiment elicited by perceived realism is a literal translation of
‘stepping into’ another person’s shoes and thus, presence may be the key to
understanding the processes of embodied experiences.
There is still much to be learned about presence and scholars are still in the
process of developing and refining the conceptual framework. Although there is
general agreement that presence should be considered a combination of several
dimensions rather than a single construct, scholars have yet to agree upon a fixed
set of components and their relationship with each other. This has led to a myriad of
confusing labels and names coined by different scholars, but there seems to be
largely three dimensions that are explored when investigating the concept of
presence: self, social, and spatial. Self presence refers to the psychological state in
which the user feels that the actual self is experiencing what the virtual self is
experiencing, or a strong connection between the virtual representation of the self
and the actual self (Biocca, 1997; Lee, 2004). Social presence can be defined as the
feeling of being with another being or individual in the mediated space (Biocca,
Harms, & Burgoon, 2003). Lastly, spatial presence is the degree to which the user
feels that the mediated environment and the objects within the environment that
surrounds him or her is real to the extent that the environment responds realistically
to user inputs (Biocca, 1997; Heeter, 1992).
Another consequence of being a relatively new concept explored in a new
medium is that scholars have yet to concur on a validated and widely used measure
of presence, despite their agreement on its importance. Most studies use self-report
25
questionnaires to gauge presence, and there is a wide array of different
questionnaires that are being used by different researchers (e.g., Biocca, Kim, &
Choi, 2001; Lessiter, Freeman, Keogh, & Davidoff, 2001; Witmer & Singer, 1998)
that define and measure presence in slightly varying ways. Not only does the
scattered use of different self-reports fail to help foster validity, but participants may
also not understand if novel terms and technical jargon are used in the questionnaire
items, leading to inaccurate responses (Slater, 2004).
Although used less frequently than self-reports, there are more objective
options of measuring presence, such as gauging physiological measures like heart
rate and skin conductance (Meehan, Razzaque, Whitton, & Brooks, 2003) or
assessing behavioral responses such as walking behavior (Usoh et al., 1999), startle
response (Slater & Usoh, 1993), reflex reactions (i.e., avoiding a vehicle heading
towards you; Loomis, 1992) to see how closely participants’ responses to mediated
stimuli match those presented in the real world. Recall measures have also been
used as an inverse proxy measure of presence (Fox, Bailenson, & Binney, 2009;
Nichols, Haldane, & Wilson, 2000) under the assumption that when users are
intensely engaged in an experience, they are likely to pay less attention to and
remember less of the information given during the engaging experience due to
limited cognitive capacity. Although more objective than self-reports, these
measures may be limited in that they are not direct measures of presence, but rather,
assume that changes in physiological and behavioral responses are in some way
connected to the perception of presence. Consequently, using a combination of both
subjective and objective measures to evaluate presence would complement the
shortcomings of each measure (Bailenson et al., 2004) and both types of measures
will be applied in the current dissertation.
26
Individual Differences in Presence Perception
Presence is a subjective state. As a result, different individuals may pay
attention to totally different stimuli and may feel different levels of presence even if
they are sharing the same mediated environment. Heeter (2003) argues that the
particular sensory stimuli that an individual might notice and pay attention to may
depend on individual differences such as personality, past experiences, or the
capacity for imagination. Steuer (1992) also proposed that the capability to feel
presence would depend on the interaction between individual factors and
technological variables. Lombard and Ditton (1997) pointed out that an individual’s
willingness to temporarily forget that he or she is in a mediated environment would
influence the level of perceived presence. Waterworth and Waterworth (2006) and
Wirth et al. (2007) suggested that individual differences in openness to experience
emotional and cognitive alterations and vivid imagination would be an important
predictor of presence, and Sas and O’Hare (2003) made similar predictions,
suggesting that individual differences in imaginative skills would be a key predictor
of presence. Thus, many scholars agree that individual differences can be major
sources of variation in the perception of presence as well as attitudes and behaviors
within virtual environments.
However, empirical evidence investigating the influence of individual
factors on presence is scarce and the few studies that have explored individual
differences yield inconclusive results. One of the earliest attempts to empirically
test the effects of personality factors on presence was conducted by Witmer and
Singer (1998). They measured individual dispositions to feel presence using the
Immersive Tendency Questionnaire, which is a scale they developed to assess
individual likelihood to feel involved or engaged in common activities (e.g., How
27
frequently do you get emotionally involved – angry, sad, or happy – in the news
stories that you read or hear?) and found positive correlations between the
Immersive Tendency Questionnaire and presence measured with a separate
Presence Questionnaire. Sas and O’Hare (2003; Sas, 2004) conducted the most
extensive series of experiments investigating the role of personality factors on the
perception of presence. They measured individual differences in empathy
(willingness to share another person’s experiences; Davis’ Interpersonal Reactivity
Index, 1983), absorption (openness to experiencing experiential events that are
sensorial or imaginal; Tellegen Absorption Scale, 1982), creative imagination
(ability to generate mental representation of objects, persons, or events not
immediately presented to the senses; Barber and Wilson’s Creative Imagination
Scale, 1979), and willingness to be transported (willingness to suspend disbelief to
enjoy a mediated experience of any kind). Results confirmed that individuals who
have higher dispositions for empathy, absorption, creative imagination and
willingness to be transported experienced significantly higher levels of presence
than those lower on these traits. On the other hand, Murray, Fox, and Pettifer (2007)
presented a study measuring individual differences in absorption, dissociation
(temporary disruption in consciousness or perception of the environment; Bernstein
& Putnam, 1986), immersive tendency, and locus of control (degree to which
people feel that they are in control of a certain event; Rotter, 1954). Results
confirmed positive correlations between dissociation and locus of control with
presence, but non-significant correlations between absorption and immersive
tendencies with presence. The non-significant relationship between absorption and
presence contradicts earlier findings, but this may be due to the fact that the two
studies used different scales to measure presence. Sporadic and inconsistent use of
28
various self-report measures of presence has been a recurring problem in this field
and this is why both self-report and objective measurements of presence will be
used in the current dissertation in an effort to offset the shortcomings of each
method.
The most recent study on personality factors and presence (Wallach, Safir,
& Samana, 2010) examined five of the individual differences mentioned above –
empathy (measured with Interpersonal Reactivity Index), imagination (measured
with Betts’ Questionnaire upon Mental Imagery; Sheehan, 1967), immersive
tendencies (measured with Immersive Tendency Questionnaire), dissociation
tendencies (measured with Dissociative Experiences Scale), and locus of control
(Locus of Control Questionnaire). Using an IVE setup similar to the one used in the
current dissertation studies with a HMD, participants experienced a virtual
simulation of an airplane ride for ten minutes. Results demonstrated that empathy
and an internal locus of control were the best predictors for the sense of presence
that participants felt during the virtual plane ride.
The current dissertation aims to advance these results by exploring
individual dispositions to understand others’ experiences and the perception of
control during embodied experiences. Based on the evidence from earlier studies,
these individual differences are expected to moderate the level of perceived
presence. Presence, in turn, is expected to be the underlying mechanism driving
attitude and behavior changes following embodied experiences.
Presence Perception in Different Modalities
Presence is often discussed in the context of virtual environments, but
scholars have argued that the illusion of “being there” may also be elicited by text
(Gerrig, 1993), television (Bracken, 2005), and by narratives presented via any form
29
of media (Green, Brock, & Kaufman, 2004). In television or video, popular
mainstream films such as The Blair Witch Project and Cloverfield have adopted a
technique that shows a first-person perspective of the story under the premise that
the protagonist is recording the entire film using a handheld camcorder. The firstperson perspective presentation amplifies dramatic tension because the audience is
able to see and hear the narrative through the protagonist’s eyes and ears. While
vividly sharing ‘in the moment’ perceptual experiences, the audience feels as
though they are standing in the shoes of the protagonist, allowing them to
momentarily disregard that the sensory content is being mediated through video and
to feel a heightened sense of realism.
Heightened presence from first-person perspectives in video also improves
their learning and training. Lindgren (2009) demonstrated that when using video
technology to assist observational learning, the first-person perspective (i.e.,
recorded by mounting a camera on the head of the person performing the task) is
more effective in learning facilitation than video recorded in the third-person
perspective (i.e., recorded by setting the camera up on a tripod). Participants who
learned how to assemble a toaster based on a video taken in first-person perspective
learned the assembly process in greater detail, remembered the process better, and
performed better on inference questions based on the learned material compared to
participants who saw the assembly process in the third-person perspective. In
another similar study (Lozano, Hard, & Tversky, 2007), participants were shown a
video clip of a person assembling a piece of furniture and told to verbally describe
the process in either the first-person perspective or third-person perspective. Later,
participants were asked to assemble the same piece of furniture. Those who had
been instructed to describe the process in the first-person perspective made
30
significantly fewer errors. There is also contrasting evidence. In a study exploring
video game playing and arousal (Lim & Reeves, 2009), participants showed no
difference in physiological arousal based on first- or third-person visual perspective
and a preference of the third-person perspective when the players were able to
choose their own avatars. The author explains that this may be due to the design of
the particular video game in that players in the first-person perspective were only
able to see the environment and not their self-representation. Because they were
unable to see their own avatars, players in the first-person perspective may not have
perceived sufficient embodiment compared to players in the third-person
perspective who had clear views of their avatar.
Presence can also be perceived from text or narrative stimuli. When reading
a narrative, individuals may feel completely immersed in the text and experience
vivid mental images of settings and characters (termed narrative transportation;
Green & Brock, 2000). Similar to the effect of presence perceived in virtual
environments, greater perception of presence from narratives has been shown to
lead to attitude change with individuals demonstrating more story-consistent beliefs
and opinions than those who felt less presence (Green, Brock, & Kaufman, 2004).
Even with text narratives, individual differences moderate the perception of
presence. Individuals with cognitive styles that enjoy effortful cognitive activity felt
higher presence in text compared to those with cognitive styles that tend to avoid
cognitive effort (Green et al., 2008).
Presence perception through text or narratives is essentially equivalent to
engaging in mental simulation and has been used frequently as experimental stimuli
in embodied cognition research. Upon reading the text or narrative, readers must
cognitively simulate the hypothetical scenario in their minds to understand the story
31
and better enjoy the vicarious experience (Green, Brock, & Kaufman, 2004). The
embodied cognition research contends that the mental simulation is based on
previously constructed sensorimotor schemata, and perceptual neurons become
activated at mere simulation of the action. This is how vivid mental simulation can
seem as real to the individual as the physical experience. Studies have also
confirmed that mental simulation can be triggered by having participants read text
stimuli and the mental simulation leads to greater perceived presence. Perceived
presence during the mental simulation, in turn, has positive effects on attitudes
toward advertisements and brands (Escalas, 2004). Thus, presence is also
hypothesized to be the underlying mechanism of mental simulation. In the current
dissertation, embodied experiences presented with IVE will be compared to
embodied experiences presented with mental simulation and presence perception in
both modalities will be measured and assessed.
32
CHAPTER FIVE: JUSTIFICATION FOR DISSERTATION STUDIES
Two studies will compare embodied experiences presented through IVE or
mental simulation (MS). MS has been used as the main manipulation of interest to
test embodied cognition in numerous studies under the premise that the body is
linked with the mind. As discussed earlier, when previously formed sensorimotor
schema is activated through MS, experiments demonstrated that mere imagination
of an activity is powerful enough to change attitude and behavior of the body. The
following dissertation studies aim to compare this traditional means of embodied
experiences against embodied experiences within IVE in several aspects.
Different Modalities and Embodied Experiences
Relevant to the embodied cognition literature, mental simulation has been
compared to physical activities in various sensorimotor contexts to confirm the
connection between the brain areas that control mental simulation and actual
sensorimotor activity. One of the earliest empirical evidence (Landauer, 1962)
proved that it took participants almost the same time to either recite the alphabet out
loud or to mentally simulate the recitation. Similarly, individuals took
approximately the same time to write a short sentence either physically or mentally
(Decety & Michel, 1989) and to walk towards targets placed at different distances
either physically or mentally (Decety, Jeannerod, & Prablanc, 1989). However,
when the task becomes difficult or awkward, for instance, moving the hand to an
unfamiliar or awkward position, mental simulation takes much longer than physical
activity and the correlation between the mind and the body becomes weak (Parsons,
1994). This implies that when individuals have no previously constructed
sensorimotor schema to base mental simulations on (i.e., no prior experience) their
simulations become either inaccurate or difficult to execute.
33
Embodied experiences presented via IVEs would overcome this limitation
of mental simulation in that they offer maximal flexibility in terms of self
representation and mediated environments. This flexibility is also bolstered with
immersion which provides vivid sensory inputs, and interactivity which provides
vivid motor inputs. Thus, for experiences that are novel or unfamiliar, IVEs are
likely to provide a higher sense of presence during embodiment compared to MS.
Only a few studies to date have compared embodied experiences within IVEs to MS.
Fox (2010) compared IVEs to MS with the purpose of promoting health behaviors,
asking participants to either watch their self representation perform pushups in an
IVE or mentally simulate themselves performing pushups. There were no
differences in behavioral outcomes between the two conditions, but female
participants believed that they could perform significantly more pushups after
watching themselves in IVE compared to MS. However, a limitation of this study is
that participants in the IVE condition had no interactive control over their self
representations and were only able to passively watch their virtual selves perform
pushups. The fact that participants felt a higher sense of identification with the self
in the MS than IVE condition confirms this speculation. Groom, Bailenson, and
Nass (2009) had participants embody a self representation of either an AfricanAmerican or a White Caucasian in an IVE or mentally simulate the embodiment.
Results indicated that embodied experiences via IVE are able to influence people’s
attitudes about racial bias whereas MS failed to do so. Participants in this study
were able to move their heads and walk inside the IVE to confirm the sense of
embodiment but still had no sensorimotor task relevant to racial stereotypes, and
thus cannot be considered a direct investigation of embodied experiences.
More relevant to the current dissertation studies, I previously conducted
34
two studies to compare embodied experiences in IVE against MS. Red-green
colorblindness, a visual disability that causes inability to perceive differences
between red and green, was chosen as the context of embodiment. Only participants
with normal color vision were recruited and embodying a colorblind individual was
a novel experience for all participants. Following earlier studies, participants were
divided into High and Low empathy groups using Davis’ Interpersonal Reactivity
Index to determine whether individual differences in empathy would moderate the
effect of embodied experiences. Participants in the IVE condition wore a HMD and
saw the virtual world treated with a colorblindness filter, as seen by a colorblind
individual. In other words, participants embodied the perceptual experience of being
colorblind and were unable to discriminate red objects from green objects.
Participants in the MS condition also wore the HMD but saw the virtual world in
normal color perspective and were asked to imagine being colorblind. Participants
in both conditions used a haptic device to complete a sensorimotor task relevant to
the embodied experience (i.e., matching red and green screws into red and green
holes).
Results demonstrated that embodied experiences within IVEs have notable
effects on participants measured to have predispositions of low empathy. For those
participants, IVE enhanced the feeling of being ‘one’ with the colorblind person
they embodied compared to the MS treatment. Furthermore, they demonstrated
more favorable attitudes toward colorblind people 24 hours after the IVE
experience. Considering that the duration of the embodied experience was
approximately five minutes, the fact that the effect carried over to the real world 24
hours after the treatment is notable. A follow-up study examined behavioral
outcomes and confirmed that participants invested twice as much time and
35
produced twice as much content to help colorblind people after embodiment in IVE
compared to the participants who relied solely on imagination. Again, it is
noteworthy that only a few minutes of embodying a colorblind person’s experience
triggered participants to spend twice as much time helping colorblind people than
participants who engaged in MS, even when this helping was completely
uncompensated.
The current dissertation studies aim to replicate and advance findings from
these prior studies. A novel context was chosen again – pro-environmental behavior.
In the first study, embodied experiences in IVE and MS were compared to
determine the most effective modality to encourage pro-environmental behaviors.
Participants were introduced to the environmental issue of using non-recycled toilet
paper and how it leads to deforestation problems. All participants were then asked
to embody a logger to experience cutting down a large tree as a result of using nonrecycled toilet paper. The study explored the degree to which actual (IVE) or
imagined (MS) sensorimotor experiences can influence participants to engage in
pro-environmental behaviors following their experiences. The second study gained
a deeper insight of embodied experiences within IVE by manipulating the levels of
immersion and interactivity separately.
It should be noted that the current dissertation studies view the experience
of cutting down a large tree in either IVE or through MS as cues that will make
environmental issues relevant to deforestation salient, rather than an experience for
participants to model. All participants were primed to think about recycling by
hearing information on how using non-recycled toilet paper leads to extensive
deforestation before engaging in or imagining the tree-cutting activity. This is in
line with earlier research on embodied cognition that demonstrated the body-mind
36
connection by showing sensorimotor experiences (e.g., holding a hot drink and
experiencing physical warmth) triggering perceptions of relevant mental schemata
(e.g., perception of warmth as a personality trait). In the same way, the sensorimotor
experience of cutting down a tree after being primed with environmental issues is
expected to trigger pro-environmental attitude and behavior in individuals, rather
than being perceived as a model intended to teach actual tree-cutting activity.
Although prior literature in embodied cognition has been successful in
inducing this body-mind connection based just on mental simulation (e.g., thinking
about past experiences triggers individuals’ body orientation to lean backwards), the
novel nature of the context of tree-cutting is expected to present challenges in
initiating the body-mind link in that the mental simulation may be difficult to
execute without sufficient sensorimotor cues. In comparison, the vivid sensorimotor
cues provided by IVEs are anticipated to facilitate the process.
Processes of Embodied Experiences
Most research on embodied cognition has demonstrated the connection
between the body and the mind by focusing on the outcome of embodied cognition.
The current dissertation studies aim to advance these studies by looking deeper into
the processes of embodied experiences. Based on earlier discussions of prior studies,
the perception of presence and control are anticipated to be the underlying
mechanisms driving the effects of embodied experiences. In particular, perception
of personal control has been shown to lead to the belief that the self’s actions can
significantly improve environmental issues and good energy behavior (Cleveland,
Kalamas, & Laroche, 2005; Curtis, Simpson-Housley, & Youck, 1984; McCarty &
Shrum, 2001). Because IVEs allow active interaction between the user and the
mediated content through tangible sensorimotor experiences, they are likely to
37
make the sense of personal control more salient and amplify the effect of embodied
experiences. Thus, the first study will compare embodied experiences in IVE and
MS in terms of presence and control and explore how these variables affect proenvironmental self-efficacy and behavior. The second study will delve deeper into
the process of embodied experiences by individually manipulating the level of
immersion and interactivity to determine their effects on perceptions of presence
and control, respectively, during embodied experiences within IVE.
Furthermore, the dissertation studies will investigate the influence of
embodied experiences on actual pro-environmental behavior. Many social science
studies rely on self-reports of behavioral intent to assess changes in behavior, but
intent to behave is not the exclusive determinant of actual behavior (Sheppard,
Hartwick, & Warshaw, 1988), and inconsistencies between behavioral intention and
actual behavior are repeatedly found. For example, earlier studies on recycling
intentions and actual recycling behavior revealed that the relationship is spurious
(Davies, Foxall, & Pallister, 2002). Self-reports of actual behavior can be inaccurate,
particularly for socially desirable norms such as pro-environmental behaviors. For
example, Bowman and colleagues (1998) found in a study on recycling behaviors
that although 72% of respondents claimed to recycle, observation revealed that only
40% of them actually participated in recycling activity. Thus, it is important to
gauge actual behavior change to supplement or compare with data from self-reports.
Individual Differences in Embodied Experiences
As discussed earlier, individual differences in the capacity to feel presence
are anticipated to be a key moderator of the effect of presence. Several studies have
found that an individual’s inherent tendency to feel empathy is positively related
with the level of presence he or she feels in the virtual environment. The previously
38
conducted colorblind studies imply that individual dispositions of empathy
moderate the effect of embodied experiences on helping behavior. Ramus and
Killmer (2007) argued that pro-environmental behavior is a special type of prosocial
behavior (i.e., behavior that is performed with the intention of promoting the
welfare of an individual, group, or organization). Taken together, these results imply
that there is a relationship between individual dispositions for empathy and proenvironmental behavior, and that presence may be an underlying mechanism
driving the relationship.
Some studies provide some empirical support for this hypothesized
relationship between individual dispositions for empathy and pro-environmental
attitude and behavior in that empathy levels mediate the relationship between the
level of self-efficacy in terms of improving the environment and the intent for proenvironmental behaviors (Allen & Ferrand, 1999). Functional magnetic resonance
imaging (fMRI) studies have also shown that individuals who self-report higher
levels of empathy demonstrate different brain activity compared to those with lower
empathy (Shane, Stevens, Harenski, & Kiehl, 2009). In this study, the anterior
cingulate was identified as the section of the brain that is involved with feeling
concern for others, and when individuals high in empathy were asked to observe
another person making errors during a task, the anterior cingulate area of their brain
showed significantly more activation compared to individuals low in empathy.
These studies evidence the importance of considering individual differences when
exploring the processes and effects of embodied experiences.
Individual differences in empathy have been usually measured with Davis’
(1983) Interpersonal Reactivity Index, a 28-item scale comprised of four subscales
that measure the individual’s propensity to engage in perspective taking and fantasy
39
as well as feel empathic concern and personal distress at the plight of others.
Together, these subscales assess a comprehensive picture of how an individual
reacts and cares about others and their needs. However, the construct most relevant
to embodied experiences is likely perspective taking, which is the mental effort to
step inside another person’s shoes – essentially the definition of embodied
experiences. Thus, a more refined scale that measures individuals’ tendency to
engage in perspective taking (Gehlbach, Brinkworth, & Wang, in press) will be used
for the dissertation studies to determine how individual differences moderate the
effect of embodied experiences.
40
CHAPTER SIX: DISSERTATION STUDY ONE: COMPARISION OF
EMBODIED EXPERIENCES WITHIN IVE AND MENTAL SIMULATION
This study directly compared embodied experiences through IVE against
MS to examine the effect of different modalities of embodied experiences on proenvironmental self-efficacy and behavior. Participants were given substantive
information on the use of non-recycled paper products and deforestation problems
and were asked to either virtually or mentally embody the role of a person cutting
down a large tree with a chain saw. In the IVE condition, participants received
visual and aural sensory inputs through a HMD and input through a haptic device to
experience cutting down a tree in an immersive virtual forest. In the MS condition,
they received a detailed text narrative of those sensory inputs and were asked to
mentally simulate the experience. Pro-environmental self-efficacy was assessed
with self-report measures and pro-environmental behavior was assessed with the
number of actual paper napkins used to clean a spill by participants following
experimental treatment.
One of the main questions of interest in Study 1 was whether the vivid
experience of embodying the tree-cutter’s perspective through IVE would exert
greater influence on actual pro-environmental behavior compared to mentally
putting oneself in the tree-cutter’s shoes. The prior colorblind studies demonstrated
that embodied experiences within IVE were able to elicit twice as much helping
behavior from participants compared to those who imagined being colorblind.
Based on these results IVE is anticipated to lead to greater behavioral change
compared to MS:
H1: Participants in the IVE condition will engage in more pro-environmental
behavior than those in the MS condition.
41
IVE’s success in influencing behavior in the non-mediated real world is
likely to be related to the higher levels of perceived presence felt during the
embodied experience due to more explicit sensory inputs available (e.g., visual,
aural, haptic) compared to MS. In addition to the self-reported measure of presence
which gauges the subjective level of realism that individuals feel in a mediated
context, recall on the information given (information on the environment provided
from outside the mediated environment) during the experimental task will be
assessed to supplement the self-reported measure. Prior studies have used memory
tasks as a proxy measure of presence to complement self-reported questionnaires
for improved validity (Fox, Bailenson, & Binney, 2009; Nichols, Haldane, &
Wilson, 2000). The logic behind using recall as a proxy for presence is that when
users are intensely engaged in an experience, they are likely to pay less attention to
and remember less of the information given during the engaging experience due to
limited cognitive capacity. Thus, higher recall would imply that the participant was
less engaged in the world, and thus felt a lesser degree of presence.
H2A: Participants in the IVE condition will report higher levels of selfreported presence than those in the MS condition.
H2B: Participants in the IVE condition will demonstrate lower levels of
recall than those in the MS condition.
Because the tree-cutting task given as the embodied experience is novel to
undergraduate participants, they will not have a previously constructed
sensorimotor schema to base their mental simulation on. Based on findings from
prior studies, individuals are likely to find MS difficult in these situations and thus
feel less in control of their imagination. On the other hand, individuals experiencing
embodiment within an IVE will receive explicit haptic input and will feel more in
42
control of their actions.
H3: Participants in the IVE condition will report higher levels of control
than those in the MS condition.
Self-efficacy is an individual’s belief that his or her actions will have
meaningful influences over events that affect their lives (Bandura, 1994). Selfefficacy is thought to determine how people feel, think, become motivated, and how
they behave. Prior studies have shown that individuals who have higher levels of
perceived self-efficacy have positive correlations with pro-environmental behaviors
(Meinhold & Malkus, 2005; Tabernero & Hernandez, 2010). In other studies on
environmental behavior, perception of control over one’s actions has been shown to
lead to the belief that the self’s actions can significantly improve environmental
issues and good energy behavior (Cleveland, Kalamas, & Laroche, 2005; Curtis,
Simpson-Housley, & Youck, 1984; McCarty & Shrum, 2001). Because IVE is
expected to elicit greater perceptions of control, it is hypothesized that:
H4: Participants in the IVE condition will report higher levels of selfefficacy after the embodied experience than those in the MS condition.
Based on prior findings suggesting that more realistic experiences lead to
attitudinal and behavioral changes, presence is thought to be an underlying
mechanism of embodied experiences. Bandura (1994) also suggests that the most
effective way of creating a strong sense of self-efficacy is through actual mastery of
experiences. Then, the more realistic an embodied experience is, the more it is
expected to increase perception of self-efficacy. Thus, presence is also expected to
be positively correlated to self-efficacy in terms of improving the environment.
Because recall is an inverse measure of presence, recall is expected to be negatively
correlated with self-efficacy and pro-environmental behavior.
43
H5A: Self-reported presence will be positively correlated with self-efficacy
and pro-environmental behavior.
H5B: Recall will be negatively correlated with self-efficacy and proenvironmental behavior.
Findings from earlier studies discussed above also suggested that
perception of control is positively correlated with presence, self-efficacy, and
behavioral change. Thus, perception of control is expected to be another underlying
mechanism of embodied experiences.
H6: Control will be positively correlated with presence, self-efficacy, and
pro-environmental behavior.
Finally, individual differences in perspective taking propensity were
explored as moderators of embodied experiences. The previously conducted
colorblind studies implied that individuals with low propensities are likely to be
more influenced by embodied experiences within IVE than MS. Another relevant
study echoed the findings from the colorblind studies. Exposing consumers with
high ability to generate mental images to imagery based appeals enhanced purchase
intention, whereas the same appeal reduced purchase intention for those with lower
ability to generate mental images (Petrova & Cialdini, 2005). This implies that
individuals with low abilities for perspective taking may not yield favorable
responses when asked to engage in MS, perhaps due to the excessive cognitive load.
H7: Participants with low perspective taking propensity in the IVE condition
will demonstrate higher self-efficacy and more pro-environmental
behavior than those in the MS condition.
Methods
Sample
44
A convenience sample was obtained from the student population of a
medium-sized West Coast university. The sample (N = 47) consisted of 29 women
and 18 men aged 18 to 46 (M = 21.60, SD = 4.27). Participants self-reported their
race/ethnicity as Caucasian/White (48.9%; n = 23); African-American/Black
(14.9%; n = 7); Latino/a (10.6%; n = 5); Asian/Asian-American (17%; n = 8); and
Other/mixed nationality (8.5%; n = 4)1.
Apparatus
The HMD used was a NVIS SX111 model which presented the virtual
environment with 640 horizontal by 512 vertical pixel resolution panels for each
eye with SXGA displays. Participants’ head movements were tracked by a threeaxis orientation sensing system (Intersense SDK version 4.19 with an update rate of
180 Hz) and used to continuously update the simulated viewpoint. The system
latency, or delay between the participant’s movement and the resulting update in the
HMD, was no greater than 80 ms. Vizard 3.0 software was used to assimilate
tracking and rendering.
The computer was also equipped with the Sensable Phantom Omni haptic
device. For the purposes of this experiment, the device’s degrees of freedom were
limited to the x-, y-, and z-axes. While most of the sawing motion occurred on the
z-axes, movement in the x- and y-axes produced micro-movements that reproduced
the realistic vibrations of a chain saw. The device offers up to 0.75 pounds of
maximum force and allows participants to touch and feel objects in the virtual
world by providing mechanical resistance based on the position of the hand as it
interacts with the virtual environment. Through the tip of the haptic device,
participants had one point of contact where the tip of the chain saw touched the
virtual object.
45
Design
All participants were subject to a pretest at least 24 hours before coming
into the lab. The pretest measured individual differences in the propensity to engage
in perspective taking and self-efficacy with regard to improving environmental
issues. In the lab, a between-subjects design was employed and participants were
randomly assigned to one of two conditions, IVE (n = 24) or MS (n = 23)2.
The IVE condition presented participants with a highly immersive treecutting experience with three layers of sensory input. Figure 1 depicts the
experimental setup for the IVE condition. Wearing the HMD, participants embodied
an avatar standing inside a forest and saw the world from the avatar’s first person
perspective in stereoscopic vision with head controlled control of point of view.
Stereovision yields perception of depth within the virtual world and the head
controlled point of view allows participants to look around in the virtual world as if
they are in the real world. The HMD also offered spatialized sound that localized
sounds in the virtual environment (i.e., chain saw, birds chirping, tree falling)
according to participants’ head movements, such that if they turned their head
toward an object on the left hand side, the volume would increase in relation to the
angle between the ear and the virtual object. Finally, using the haptic device,
participants were able to feel realistic vibrations of the chain saw they used to cut
the tree with and saw the arms of the avatar move in sync with movement of their
physical arms. To heighten the realism of the experience, a physical handle that
resembles the handle of the chain saw used in the IVE was constructed out of
cardboard and attached to the haptic device. Details such as birds flying in and out
of the trees, sound of running water, and grassy terrain made the virtual forest
realistic.
46
The MS condition presented participants with a narrative description of this
virtual forest. To create maximum similarity between the MS condition and the rich
sensory inputs offered in the IVE condition, 14 judges from a separate sample from
the main study (7 males, 7 females) were asked to pretest the IVE condition to help
develop the text stimulus. Two methods were combined to capture all the sensory
description that the judges could come up with regarding the virtual forest. First, as
the judges experienced the virtual forest and the tree-cutting procedure, they were
asked to think out loud, taking care to describe every sensory detail in the virtual
forest, including everything that they saw, heard, and felt. These real-time thoughts
were recorded with an audio device and later transcribed. Secondly, after the judges
finished cutting the tree down in the virtual forest, they were subject to a thoughtlisting procedure (Cacioppo & Petty, 1981), which asked participants to write down
all of their thoughts during the augmented experience. Based on the information
from transcribed audio files and the thought-listing procedures, a narrative stimulus
that included all of the verbal information gathered from the 14 judges was
developed for the MS condition (Appendix A). Because all of the judges
experienced the same virtual forest, the sequence of events that occurred in the
forest was the same for all of them (i.e., looking around the forest, sawing the tree,
watching the tree fall). To develop the final MS stimulus, information that was
repeated or redundant among the judges was included only once in the narrative and
grammatical errors were corrected.
47
Figure 1. Experimental Setup for the IVE Condition in Study 1 (A).
Participants wore the HMD (C) and were able to look around the virtual forest in
stereovision and heard spatialized aural inputs wearing headphones (headphones not
shown in figure). The HMD (C) was equipped with an Intersense cube that allowed
head-controlled point of view in the IVE. They were then instructed to pull and
push the haptic device (B) to saw the virtual tree down.
Procedure
Upon arrival at the lab, participants in the IVE condition first received
instructions on the haptic device and were shown how to operate it. Then they were
told that they would embody a “tree-cutter” in the virtual world who is about to cut
down a tree. Wearing the HMD, participants entered the virtual world and saw the
forest through the eyes of the tree-cutter standing in front of a tree holding onto the
handle of an chain saw, poised to begin cutting the tree down in front of him or her.
Figure 2 displays the series of events that the participant experiences in the IVE.
48
Before engaging in any cutting activities, participants were first asked to
look around the forest, taking note of details such as birds, plants, sky, ground, and
the body of the tree-cutter they were embodying. The experimenter then read out
loud a short piece of information that revealed how many rolls of non-recycled
toilet paper a single cord of tree is able to produce, and how many trees had been
cut down to supply an average 20 year-old American with non-recycled toilet paper.
Participants then heard the sound of the chain saw starting, and were instructed to
begin moving the haptic device back and forth to cut the tree down. The program
required all participants to engage in cutting motions for 2 minutes. After 2 minutes,
participants saw and heard the tree trunk crash down to the ground and were asked
to look around the forest once more. As a result of cutting the tree down, the forest
was programmed to become completely quiet and all movement was removed to
emphasize the damage inflicted upon the forest. Participants were then guided
toward a survey computer where they filled out a series of questionnaires regarding
their experience as a tree-cutter and the environmental information they recalled.
For the MS condition, participants were told that they would mentally take
the perspective of a “tree-cutter” who is about to cut down a tree. Instructions
directed participants to try to put themselves in the shoes of the tree-cutter, imagine
his or her perspective, and to think of the tree-cutter as an extension of himself or
herself. They were also asked to create a vivid picture in their minds about what
they see, hear, and feel in the forest while reading the narrative stimulus.
Participants were then given the narrative stimulus that contains detailed
descriptions of all the sensory experiences from the IVE condition. The narrative
stimulus was designed to have a break in the middle where the participants in the
IVE condition would receive information about using non-recycled toilet paper and
49
deforestation issues. When participants finished reading the first part of the passage,
the experimenter read out the same piece of information that was given in the IVE
condition, and then instructed participants to finish reading the rest of the passage.
Thus, participants in both conditions received pro-environmental information at
approximately the same point during the experimental treatments (i.e., after actually
looking around the forest in the IVE condition; after reading the descriptions of the
forest in the MS condition). When they finished reading, participants were guided
toward a survey computer where they filled out a series of questionnaires regarding
their experience as a tree-cutter and the environmental information they recalled.
Finally, the last procedure of the experiment involved assessing
participants’ pro-environmental behavior. In order to allow some time to pass after
the experimental treatment so that the immediate sensitization toward
environmental issues after experimental treatments would wear off, participants
were asked to participate in a 30-minute long additional experiment that was
completely irrelevant to the first one. Upon completion of this second experiment,
participants were seated at a table that had a cup holding some water, and asked to
fill out a demographic information sheet. The water was pre-measured to ensure that
all participants received the same amount of water, using the same cup each time.
While participants were busy filling out the form, the experimenter approached the
table to gather some forms for the next participant and in the process, knocked over
the cup of water. The experimenter then asked each participant for help by saying,
“I’m so sorry, but I have to prepare the next participant for the experiment. Could
you help me clean the water up?” and handed a pre-counted number of paper
napkins to the participant. After the participant left, the number of used napkins was
counted and recorded as an inverse measure of pro-environmental behavior.
50
The number of napkins that participants use would be an effective proxy for
measuring pro-environmental behavior because the experimenter would be able to
observe actual changes in the participant’s usage of paper products rather than
relying on a self-report measure. It should also be noted that this behavior was
assessed after all experimental treatment and surveys were completed and placed
unsuspecting participants in a situation that required immediate response (i.e.,
cleaning up spilled water with napkins) without giving them time to think about
socially desirable or ‘correct’ responses. This way, the experimenter was able to
observe naturalistic responses outside the constraints of an experimental context.
Figure 2. Screenshot of the Series of Events in the IVE Condition in Study 1.
Participants saw the virtual forest in first-person perspective and were first
instructed to take a good look around (A). After hearing some information on using
non-recycled toilet paper and the resulting problems of deforestation, participants
were instructed to begin sawing down the large tree in front of them (B). Using the
haptic device, participants moved the chain saw in and out for two minutes (C) until
the tree finally fell, crashing to the ground (D).
51
Measures
Items for all measures included in this study can be viewed in Appendix B.
Perspective Taking Propensity. This measure was administered in the
pretest. Seven items from Gehlbach’s Social Perspective Taking Propensity Scale
(in press) assessed each individual’s disposition to try to take the perspective of
another person. Participants were asked to indicate on a 5-point scale (1 = Almost
never; 5 = Almost all the time) how often they attempted to understand and try to
put themselves in the shoes of another (e.g., “Overall, how often do you try to
understand the point of view of other people?). Results were averaged; scores
ranged from 1.86 to 5.00 (M = 3.65, SD = .70). Reliability for all seven items was
Cronbach’s α = .88. As in the colorblind studies, a median split was conducted to
split the participants into High Propensity (n = 25) and Low Propensity (n = 21)
groups.
Napkins. The number of napkins used to wipe the water off the table was
counted as an inverse measure of pro-environmental behavior. The number of
napkins used ranged from 0 to 12 (M = 5.11, SD = 2.39).
Presence. Eleven items were used to assess participants’ experience of
presence while immersed in the virtual environment or in the mentally simulated
world. These items were culled from several sources (Bailenson & Yee, 2007;
Nowak & Biocca, 2003; Witmer & Singer, 1998). Participants indicated on a 5point scale (1 = Not at all; 5 = Extremely) the degree to which they felt that they
had embodied the tree-cutter and the extent to which they felt the forest and the
tree-cutting experience were real. Responses were averaged; scores ranged from
1.17 to 4.42 (M = 2.79, SD = .84). Reliability for this measure was Cronbach’s α
= .94.
52
Recall. Using the thought-listing procedure, participants were asked to
remember and write down all the pieces of information on toilet paper production
and its effects on the forest. Each correctly recalled information was given a point.
Two raters blind to experimental conditions counted the number of correctly
recalled information. Scores ranged from 0 to 3, and Cohen’s κ was .81, indicating
high inter-coder reliability (Landis & Koch, 1977). Coded results from the two
raters were averaged to establish a single measure of recall.
Control. The Self Assessment Mannekin (SAM; Lang, 1980) assessed the
level of control felt by participants during the tree-cutting process. SAM is a
nonverbal, pictorial assessment technique and the semantic differential rating
system is a verbal rating system to measure a person’s reaction to a stimulus. The
SAM is a single pictorial item that depicts the increasing level of felt control by
changes in the size of SAM (i.e., larger SAM denotes greater felt control) on a 9point scale (1 = small SAM ; 9 = large SAM). Bradley and Lang (1994) found that
SAM can be easier to administer than semantic differential scales because it is a
more direct measure that only requires simple judgment of explicit pictures. They
also found that SAM is more accurate than semantic differential scales in terms of
measuring participants’ perception of control over the stimuli.
Self-Efficacy. This measure was administered once in the pretest and again
after experimental treatments to capture the effect of respective treatments on selfefficacy toward environmental issues. Ten items from the Environmental Action
Internal Control Index (Smith-Sebasto & Fortner, 1994) assessed participants’
perception of self-efficacy with regard to improving environmental issues.
Participants answered on a 5-point scale (1 = Does not describe my point of view
well; 5 = Describes my point of view very well) the extent to which they agreed to
53
statements that describe how individual actions can improve the environment (e.g.,
“My individual actions would improve the quality of the environment if I were to
buy and use recycled paper products”). Reliability for this measure was high with
Cronbach’s α = .94. Because pre-treatment self-efficacy was highly correlated with
some of the dependent variables, it was entered as a covariate for subsequent
analyses to control for individual differences prior to the experimental treatment.
Videogame. Participants were asked to write the average time they spent
playing video games each week. Familiarity with virtual environments developed
by extensive exposure to video games could influence subsequent responses to the
IVE condition. Earlier studies have shown that prior gaming experiences change the
way that individuals respond to and navigate in virtual environments (Jelfs &
Whitelock, 2001). Also, experience playing action video games has been shown to
eliminate sex differences in spatial attention and decrease sex differences in mental
rotation ability (Feng, Spence, & Pratt, 2007). To control for these individual
differences that may lead to unwanted variance, this measure was entered as a
covariate in subsequent analyses.
Results
Means and standard deviations for all dependent variables can be viewed in
Table 1. Appendix D lists statistics for all significant and non-significant results
from the analyses.
54
Table 1
Study 1 Means and Standard Deviations for Dependent Variables by Condition
Napkins
DV
Control
Presence
Recall
Posttreatment
Self-efficacy
Condition
IVE
M
SD
M
SD
M
SD
M
SD
M
SD
4.61
2.35
5.22
1.93
2.80
.84
.75
.68
3.38
1.01
MS
5.61
2.43
3.70
2.22
2.66
.89
1.43
.83
3.70
.93
Napkins
To gauge the level of actual pro-environmental behavior elicited by
embodied experiences, an ANOVA was run with the number of used napkins as the
dependent variable, experimental conditions and perspective taking propensity as
the independent variables, and videogame hours and pre-treatment self-efficacy as
covariates. The main effect of experimental condition was significant, F(1, 40) =
4.16, p < .05, partial η2 = .09, with participants in the IVE condition using
significantly less napkins than those in the MS condition. H1 was supported. The
main effect of perspective taking propensity was also significant, F(1, 40) = 4.73, p
< .05, partial η2 = .11, with participants in the High Propensity group using
significantly less napkins (M = 4.48, SD = 1.94) than those in the Low Propensity
group (M = 6.75, SD = 3.41). No interaction effects were significant, and H7 was
not supported.
Presence and Recall
An ANOVA was run with presence as the dependent variable, and the same
independent variables and covariate as above. No main or interaction effects were
significant, all Fs < 1. Because no significant differences in the level of presence
were found between the two conditions, H2A was not supported.
55
Another ANOVA was performed with recall as the dependent variable, and
the same independent variables and covariate as above. The main effect of
experimental condition was significant, F(1, 40) = 6.22, p < .05, partial η2 = .14,
with the IVE condition yielding significantly lower recall scores compared to the
MS condition. H2B was supported. No other main or interaction effects were
significant. This yields partial evidence that participants in the IVE condition felt
higher levels of presence and were more engaged in the mediated environment,
resulting in lower recall scores compared to those in the MS condition.
Control
Next, an ANOVA was performed with the single SAM item for personal
control as the dependent variable, and the same independent variables and covariate
as above. The main effect of experimental condition was significant, F(1, 40) = 4.44,
p < .05, partial η2 = .10, with the IVE condition yielding perceptions of greater
control compared to the MS condition. No other main or interaction effects were
significant, all Fs < 1. H3 was supported.
Self-Efficacy
To compare the change in self-efficacy levels before and after the
experimental treatments, a repeated-measures ANOVA was run with both selfefficacy measures as the dependent within-participants variable, and the same
independent variables and covariate above as the between-participants variables.
The main effect for self-efficacy was significant, F(1, 41) = 4.62, p < .05, partial η2
= .10, with significantly higher self-efficacy levels after the experimental treatments
(M = 3.56, SD = .97) compared to before the treatments (M = 3.13, SD = 1.00). No
other effects were significant. Because only the main effect of self-efficacy was
significant, H4 was not supported and H7 were not supported.
56
Correlations among Dependent Variables
Pearson correlations among all dependent variables are shown in Table 2 to
take a closer look at relationships among them.
Table 2
Pearson Correlations between Dependent Variables in Study 1 (N = 47)
DV
Napkins
Presence
Recall
Control
Pre. SE
Post. SE
Napkins
--
.14
-.04
.11
-.16
-.10
Presence
.14
--
-.17
.19
.11
.28
Recall
-.04
-.17
--
-.42**
.04
.05
Control
.11
.19
-.42**
--
-.12
-.22
-.16
.11
.04
-.12
--
.65**
-.10
.28
.05
-.22
.65**
--
Pre-treat
SE
Post-treat
SE
**Correlation is significant at p < .01.
Contrary to expectations, self-reported presence and recall was not
positively correlated with self-efficacy or pro-environmental behavior (i.e., napkins),
and H5A and H5B were not supported. Control was also not positively correlated
with self-efficacy or pro-environmental behavior, but was negatively correlated with
recall, yielding partial support for H6. This result echoes the findings from Sas
(2004) and Wallach, Safir, and Samana (2010) in which perception of control was
positively correlated with the level of presence felt in the virtual environment.
Finally, pre- and post-treatment self-efficacy levels demonstrated a strong
57
positive linear relationship. This implies that although there was a significant
increase of self-efficacy following experimental treatment, the relative difference in
self-efficacy levels between individuals were maintained over time, before and after
the embodied experience. This adds support to the importance of inherent individual
differences as variables in embodied experiences.
Discussion
This study yielded several important insights regarding embodied
experiences presented through IVE and MS and how they influence self-efficacy
and behavior. After experiencing rich sensory experiences of the tree-cutter in the
IVE or by mentally simulating the same experience, participants in both conditions
demonstrated a significant increase in self-efficacy. That is, all participants,
regardless of the modality of their embodied experiences felt a higher sense that
their individual actions could make significant changes to improving the
environment compared to before their embodied experiences.
However, differences between experimental conditions were manifested
when it came to actual pro-environmental behavior. Results indicated that
participants who embodied the tree-cutter in an IVE used significantly less napkins
compared to participants who imagined the embodiment. Thus, it seems that
increasing awareness of environmental issues through either embodied experiences
within IVE or MS fosters a greater sense of self-efficacy. However, increase in selfefficacy does not always lead to actual pro-environmental behavior, as
demonstrated by the napkin usage. The results of this study imply that actual
sensorimotor experiences within IVEs (i.e., visual, aural, and haptic) are more
powerful than mere imagination in terms of transferring over to the physical world
to influence actual behavior.
58
The hypothesis that embodied experiences within IVEs would result in a
higher perception of presence compared to MS was not supported by the self-report
questionnaires but indirectly demonstrated with the recall measure. This partially
implies that embodied experiences within IVEs seem more realistic to users
compared to MS. In an effort to improve the inconsistent results between the
subjective and objective measures of presence, the second dissertation study will
attempt to gather richer subjective data on presence by using an enhanced presence
questionnaire along with the recall measure. The inconsistency may also be
explained by the fact that the presence scale used for this study was mostly intended
for use after exposure to virtual environments. Using a presence scale specifically
developed for narratives (e.g., Green & Brock, 2000) may have been a more
sensitive measure for the MS condition. For instance, rather than asking “How
much did the forest seem real?” “How much could you picture yourself in the scene
of events described in the narrative?” may have been a more appropriate item to
measure presence in the MS condition.
The perception of control was significantly affected by the type of modality,
with participants in the IVE feeling greater levels of control compared to
participants in the MS condition. This result implies that individuals may not feel in
control of their embodied actions when they have no previously constructed
sensorimotor schema to base their mental simulation on.
Contrary to expectations, presence and control were not positively
correlated to self-efficacy and pro-environmental behavior. This may be due to the
fact that the self-reported presence measure used in this study was not sensitive
enough to pick up felt level of realism. Earlier studies have demonstrated that
indirect and objective behavioral measures can better assess presence (Bailenson et
59
al., 2004). In terms of control, high interactivity may not always be the optimal
solution to increase self-efficacy and favorable behavior (Sundar, 2000). Thus,
Study 2 will manipulate different levels of immersion and control to obtain deeper
insight into the mechanism of embodied experiences in IVEs.
Finally, individual differences in perspective taking propensity did not
moderate self-efficacy or pro-environmental behavior. However, participants with
High Propensity did use significantly less napkins than those with Low Propensity,
indicating that individual differences are involved in embodied experiences to some
extent. This variable will also be further explored in the following study.
60
CHAPTER SEVEN: DISSERTATION STUDY TWO: EFFECT OF IMMERSION
AND AGENCY OF CONTROL ON EMBODIED EXPERIENCES
Study 2 set out to answer lingering questions on the processes of embodied
experiences. Although results demonstrated that embodied experiences in IVE
elicited higher recall, control, and more pro-environmental behavior than MS, it is
still unclear exactly which factors of IVE led to those results. Thus, Study 2 will
parse out different elements of embodied experiences in IVE, manipulating the
elements of immersion and control individually to investigate its effect on selfefficacy and behavior.
As previously discussed, the level of immersion is most closely related to
the level of perceived presence. Although there is no standardized quantification
used to divide levels of immersion, by definition, more layers of simulated
sensorimotor inputs mean greater immersion. In Study 2, immersion will be varied
in two levels – High Immersion (stereoscopic vision, spatialized sound, head
controlled point of view, and naturalistic tactile feedback) or Low Immersion
(monoscopic vision, non-spatialized sound, manually controlled point of view, and
abstract tactile feedback).
Control will also be varied in two levels by manipulating the agency of
control – Self-Move (the participant is the entity in control of moving the saw to cut
the tree down) or Other-Move (the participant follows the movement made by
another participant to move the saw to cut the tree down). Presence and control
failed to correlate with self-efficacy and pro-environment behavior in Study 1, but
by manipulating them as independent variables, Study 2 may yield deeper insight
into the processes of embodied experiences presented within IVEs.
Results of Study 1 demonstrated that participants felt higher presence (as
61
measured by recall) during their embodied experience in the IVE, felt higher control
in the IVE condition, and that the IVE condition ultimately yielded greater proenvironmental behavior. Other studies found that prior personal experience of
recycling behavior have been found to be strong predictors of future recycling
behavior (Bagozzi & Dabholkar, 1994; Boldero, 1995). Then, embodied
experiences within IVEs may be sufficiently realistic and analogous to building
personal experiences and the relevant sensorimotor schemata. Consequently, it is
anticipated that the more immersive the embodied experience is and greater the
perception of control, the greater the influence will be on pro-environmental
behavior by effectively constructing a sensorimotor schema to store in the
individual’s memory.
H8: Participants in the High Immersion, Self-Move condition will
demonstrate the greatest actual pro-environmental behavior compared to
all other conditions.
As discussed earlier, intent to behave does not necessarily lead to actual
behavior, particularly when the behavior is considered socially desirable. Thus, selfreported intentions to engage in pro-environmental behavior will be compared to
actually manifested behavior to supplement the results from Study 1. Comparing
both types of measurements may yield further insight to how embodied experiences
can lead to behavioral change.
RQ1: How will intention to engage in pro-environmental behavior compare
to actual pro-environmental behavior?
Following Study 1, recall of information given during the perspective taking
experience will be assessed again as a proxy measure of presence to supplement
self-reported measures of presence. Based on the results from Study 1, high level of
62
immersion is hypothesized to yield high levels of recall. Also, Study 1 demonstrated
that recall was negatively correlated with control. Thus, it is hypothesized that:
H9A: Recall will be lower in the High Immersion condition compared to the
Low Immersion condition.
H9B: Recall will be lower in the Self Move condition compared to the Other
Move condition.
Also based on the results of Study 1, an overall increase in self-efficacy
following the embodied experience, regardless of experimental condition, is
expected.
H10: Pre-treatment self-efficacy will significantly increase following any
embodied experience.
In addition to manipulating the agency of control, Study 2 will examine
locus of control to assess how having self- or other-controlled haptic feedback
during embodied experiences affect the degree to which participants reflect upon
themselves for the environmental damage caused (i.e., cutting the tree down). Locus
of control, as originally defined by Rotter (1954), refers to the extent that
individuals internally attribute the cause of events that affect them. Numerous
studies show that an internal locus of control is the driving factor leading to proenvironmental attitude and behavior (Balderjahn, 1988; Curtis, Simpson-Housley,
& Youck, 1984; Henion & Wilson, 1976; Shrum, Lowrey, & McCarty, 1994). Using
IVEs, researchers now have the freedom to manipulate the agency of control to
influence the locus of control which is something that was not possible with
traditional media3. Varying the level of interactivity that participants are exposed to
will yield further insights on the role that control plays in embodied experiences.
In the Self-Move condition, participants had active control of the haptic
63
device synced with the movement of the virtual saw cutting down the tree. In the
Other-Move condition, the computer was programmed to play back movements
from another participant. Thus, participants in the Other-Move condition embodied
the tree-cutting avatar but passively experienced the tree-cutting by letting the
haptic device guide his or her physical arms through the sawing movements. It is
likely that active agents of control will place greater internal attributions about
cutting the tree down (as a result of using non-recycled toilet paper) than passive
agents of control.
H11: Participants in the Self-Move condition will make more internal
attributions for cutting the tree down compared to participants in the
Other-Move condition.
Finally, individual differences in perspective taking propensity were
explored again to see how they moderate the effect of different levels of immersion
and agency of control during embodied experiences.
RQ2: How will individual differences in perspective taking propensity
moderate the effect of immersion and agency of control on self-efficacy
and pro-environmental behavior?
Methods
Sample
A convenience sample was obtained from the student population of a
medium-sized West Coast university. The final sample (N = 101) consisted of 60
women and 41 men aged 17 to 39 (M = 20.78, SD = 2.65)4. Participants selfreported their race/ethnicity as Caucasian/White (42.6%; n = 43); AfricanAmerican/Black (5.9%; n = 6); Latino/a (6.9%; n = 7); Asian/Asian-American
(28.7%; n = 29); and Other/mixed nationality (15.8%; n = 16).
64
Apparatus
The same experimental set up from Study 1 was used. The Sensable
Phantom Omni haptic device was replaced with the Novint Falcon with three
degrees of freedom (x, y, and z), providing up to two pounds of force feedback. By
increasing the force in feedback, participants were able to receive more realistic
tactile input of the chain saw.
Design
All participants were subject to the pretest used in Study 1 at least 24 hours
before the experimental treatment, assessing their level of perspective taking
propensity and pre-treatment self-efficacy.
In the lab, participants were randomly assigned to a condition following a
2x2 between-participants design. One independent variable was Agency of Control,
separated into Self-Move and Other-Move, which manipulated the person in control
of the saw movements. The second independent variable was Immersion, separated
into High and Low Immersion, which manipulated the number and range of sensory
inputs that the participant received while exposed to the virtual environment.
Table 3 describes the operationalization of each cell in greater detail.
65
Table 3. 2x2 Between-Participants Design for Study 2
Other-Move
Self-Move
Low Immersion (Desktop)
Participant views the virtual world using a desktop monoscopic monitor without spatialized sound. Participant enters the virtual world in first person perspective. Participant manually controls the point of view using a mouse. Participant has no control of the saw. High Immersion (HMD)
Participant wears a HMD with stereoscopic display and spatialized sound. Participant enters the virtual world in first person perspective. Participant has head controlled point of view. Participant has no control of the saw, but is guided by the movement made by another participant through holding the haptic device. Participant views the virtual Participant wears a HMD with stereoscopic screens and world using a desktop spatialized sound. Participant monoscopic monitor without enters the virtual world in first spatialized sound. Participant person perspective. Participant enters the virtual world in first has head controlled point of view. person perspective. Participant manually controls the point of Participant controls the saw, view using a mouse. making physical sawing Participant controls the saw with movements with the haptic device to cut the tree down. a button. Each button press moves the saw once. To control for variance in the tree-cutting movement that participants
experience in the Other-Move condition, tracked information from the participant in
the Self-Move condition directly preceding the Other-Move condition was used as
the stimulus presented in the Other-Move condition. That is, the saw movement
made by participants in the Low Immersion/Self-Move condition and High
Immersion/Self-Move condition was recorded and played back for participants
watching the tree being cut down in the succeeding Low Immersion/Other-Move
condition and High Immersion/Other-Move conditions, respectively.
Yoking participants between these conditions was planned in advance in the
experimental design to ensure that the tracked information for each participant in
66
the Self-Move condition was used once for each participant in the Other-Move
condition. Consequently, across the two conditions the movements experienced as a
whole were quite similar.
Procedures
Low Immersion/Self-Move condition. Upon arrival at the lab, participants
received instructions on the haptic device and were shown how to operate it. Then
they were told that they would gain personal experience of cutting a tree down by
embodying the body of a tree-cutter who is about to cut down a tree. At a desktop
computer station, participants saw the virtual forest used in Study 1 (monoscopic
vision). Participants were able to move their point of view manually using a mouse
and were given headphones to wear to receive aural inputs. Figure 3 depicts the
experimental setting for the Low Immersive/Self-Move condition.
Figure 3. Experimental Setup for Low Immersion, Self-Move Condition in Study 2.
Participants wore headphones for non-spatialized aural input, and pressed the
spacebar to control the chain saw.
67
Before engaging in any cutting activities, participants were first asked to
look around the forest using the mouse, taking note of details such as birds, plants,
sky, ground, and the body of the tree-cutter they were embodying. While looking
around in the world, participants heard a pre-recorded audio file narrating the
information on toilet paper usage and deforestation used in Study 1. Afterwards,
participants heard the sound of the chain saw starting and were instructed to hit the
spacebar to move the saw to cut the tree down. Each spacebar press moved the saw
in and out once. All other procedures were same as Study 1. After spending 2
minutes cutting down the tree, participants were guided toward a survey computer
where they filled out a series of questionnaires regarding their experience as a treecutter.
Low Immersion/Other-Move condition. Participants experienced the same
virtual forest under the same experimental setup as the Low Immersion/Self-Move
condition. However, these participants had no control of the saw. Rather, the
movements of the saw made by the participant in the preceding Low
Immersion/Self-Move condition were tracked and played back such that the haptic
device pulled and pushed the participants’ physical arms. In essence, participants in
the Low Immersion/Other-Move condition did not actively cut the tree down but
watched the recorded movie of another participant cutting the tree down for 2
minutes.
High Immersion/ Self-Move condition. Participants experienced the virtual
forest with a rich array of sensory information. Participants wore the HMD and saw
the forest in stereoscopic vision with spatialized audio and had head-controlled
point of view. They were also given full control of the haptic device to cut the tree
down. As in Study 1, the physical movement of the haptic device made by the
68
participant was synchronized to the movement of the virtual arms of the tree-cutter.
Participants engaged in active tree-cutting movements for 2 minutes.
High Immersion/Other-Move condition. Participants experienced the virtual
forest in stereoscopic vision and spatialized sound and were given head-controlled
point of view, but had no control of the saw. In this condition, participants held on
to the handles on the device while the computer guided the movement of their hands.
Similar to the Low Immersion conditions, the haptic device’s movement was a
playback of the movements the participant in the preceding High Immersion/SelfMove condition made.
Measures
All specific survey items can be found in Appendix C.
Perspective Taking Propensity. The same measure from Study 1 was used.
Reliability for all seven items was Cronbach’s α = .81. Scores ranged from 2.29 to
5.00. A median split was conducted to split the participants into High Propensity (n
= 49) and Low Propensity (n = 52) groups.
Manipulation Check for Agency of Control (SAM). The SAM used in Study
1 assessed the successful manipulation of Self-Move and Other-Move conditions.
Scores ranged from 1 to 9 (M = 4.83, SD = 2.15).
Manipulation Check for Immersion (Presence). A refined measure of
presence was administered to check the successful manipulation of High and Low
Immersion conditions. Six items from Ratan’s (2010) Self Presence Questionnaire
and 11 items from the ITC-Sense of Presence Inventory (Lessiter, Freeman, Keogh,
& Davidoff, 2001) were adapted to fit the current study. Reliability for this measure
was Cronbach’s α = .93. Responses were averaged; scores ranged from 1.00 to 4.06
(M = 2.78, SD = .75).
69
Napkins. The same behavioral measure from Study 1 was used. The number
of napkins used ranged from 0 to 14 (M = 8.01, SD = 3.33).
Behavioral Intention. Five items were created to assess participants’
intentions to engage in recycling behavior. Five-point, fully labeled Likert scales
measured the extent to which they planned to engage in recycling. Reliability for
this measure was Cronbach’s α = .80. Responses were averaged; scores ranged from
1.00 to 4.20 (M = 2.70, SD = .73).
Recall. The recall measure from Study 1 was used. Scores ranged from 0 to
8 (M = 3.47, SD = 1.53), and Cohen’s κ was .70, indicating good inter-coder
reliability (Landis & Koch, 1977). Coded results from the two raters were averaged
to establish a single measure of recall.
Self-Efficacy. The measure from Study 1 was used again for pre- and posttreatment assessments. Reliability for this measure was high with Cronbach’s α
= .93 for both the pre- and the post-treatment self efficacy items. As in Study 1, this
measure was entered as a covariate in subsequent analyses to control for individual
and sex differences.
Locus of Control. Three items from the Revised Casual Dimension Scale
(CDSII; McAuley, Duncan, & Russell, 1992) were used. Participants were first
asked to respond an open-ended question about the reason they (i.e., through the
tree-cutter) cut the tree down in the virtual forest. The scale items then asked them
to think about those reasons and to indicate the extent to which they agreed to 9point semantic differential scales that measure the locus of control (1 = Internal
locus of control; 9 = External locus of control). Specific items can be found in
Appendix C. Reliability for this measure was Cronbach’s α = .76. Responses were
averaged; scores ranged from 1.00 to 6.33 (M = 2.98, SD = 1.30).
70
Videogame. As in Study 1, average weekly hours of videogame play was
entered as a covariate to control for individual variances.
Results
Means and standard deviations for all dependent variables can be viewed in
Appendix E.
Manipulation Checks
An ANOVA was run with SAM as the dependent variable, agency of
control, immersion, and perspective taking propensity as the independent variables,
and videogame hours and pre-treatment self-efficacy as covariates. The main effect
of agency of control was significant, F(1, 91) = 6.30, p < .05, partial η2 = .07,
confirming the success of manipulation between the Self-Move and Other-Move
conditions. All other effects were non-significant.
Another ANOVA was run with presence scores as the dependent variable
and the same independent variables and covariates. The main effect of immersion
was significant, F(1, 91) = 9.59, p < .01, partial η2 = .10, confirming the success of
manipulation between the High and Low Immersion conditions. All other effects
were non-significant.
Behavioral Intention and Napkin Use
An ANOVA was run with behavioral intention as the dependent variable
and the same independent variables and covariates as above. The main effect of
perspective taking propensity was significant, F(1, 91) = 4.52, p < .05, partial η2
= .05, with participants with High Propensity reporting significantly higher
intentions to recycle (M = 2.74, SD = .66) than participants with Low Propensity (M
= 2.67, SD = .79). No other effects were significant.
Next, the same ANOVA was run with actual behavior (i.e., napkins) as the
71
dependent variable. As shown in Figure 4, the interaction between agency of control
and perspective taking propensity was significant, F(1, 91) = 4.86, p < .05, partial
η2 = .05. Posthoc analyses with Tukey’s HSD revealed that participants with Low
Propensity used significantly less napkins when they experienced the Other-Move
condition (M = 7.30, SD = 3.29) than the Self-Move condition (M = 8.60, SD =
3.54), whereas participants with High Propensity used significantly less napkins
when they experienced the Self-Move condition (M = 7.35, SD = 3.64) than the
Other-Move condition (M = 8.96, SD = 2.55). All other main and interaction effects
were non-significant, and H8 was not supported.
Figure 4. Interaction between Agency of Control and Perspective Taking Propensity
on Napkin Use in Study 2
Recall
An ANOVA was run with recall as the dependent variable and the same
independent variables and covariates. The main effect of immersion was significant,
F(1, 91) = 6.24, p < .05, partial η2 = .06, with participants in the High Immersion
72
condition recalling significantly less information (M = 3.13, SD = 1.42) compared
to participants in the Low Immersion condition (M = 3.81, SD = 1.58). Also, selfreported presence measures from the manipulation check above confirmed that
High Immersion conditions are higher in presence than Low Immersion conditions
and this result further confirms that recall is an effective proxy measure of selfreported presence. H9A was supported. However, the main effect of agency of
control was not significant, and H9B was not supported.
Self-Efficacy
To compare the change in self-efficacy levels before and after experimental
treatments, a repeated-measures ANOVA was run with both self-efficacy measures
as the dependent within-participants variable, and the same independent variables
and videogame hours as the covariate. The main effect for self-efficacy was
significant, F(1, 92) = 41.93, p < .01, partial η2 = .31, with significantly higher selfefficacy levels after the experimental treatments (M = 3.65, SD = .79) compared to
before the treatments (M = 3.22, SD = .75), regardless of experimental condition.
H10 was supported.
In addition, the interaction between self-efficacy and perspective taking
propensity was also significant as shown in Figure 5, F(1, 92) = 8.91, p < .01,
partial η2 = .09.
73
Figure 5. Interaction between Self-Efficacy and Perspective Taking Propensity Over
Time in Study 2
Posthoc analyses with Tukey’s HSD revealed that for participants with
High Propensity for perspective taking, self-efficacy became significantly higher
post-treatment (M = 3.78, SD = .68) than pre-treatment (M = 3.52, SD = .54). Also
for participants with Low Propensity for perspective taking, self-efficacy became
significant higher post-treatment (M = 3.53, SD = .87) than pre-treatment (M = 2.95,
SD = .82). The post-treatment self-efficacy score for the High Propensity group was
significantly higher than all other conditions. Posthoc tests also revealed that
embodied experiences were more effective for participants with Low Propensity
than those with High Propensity in that individuals with Low Propensity
experienced twice as much increase in self-efficacy (.26 mean score difference)
following the embodied experience compared to those with High Propensity (.58
mean score difference).
74
Locus of Control
An ANOVA was run with locus of control as the dependent variable and the
same independent variables and covariates. The main effect of agency of control
was significant, F(1, 91) = 4.73, p < .05, partial η2 = .05, with participants in the
Self-Move condition making more internal attributions for the environmental
damage (i.e., cutting the tree down) (M = 2.70, SD = 1.22) compared to participants
in the Other-Move condition (M = 3.26, SD = 1.32). H11 was supported.
Correlations Among Dependent Variables
Once again, Pearson correlations among all dependent variables are shown
in Table 4 to take a closer look at relationships among them.
Table 4
Pearson Correlations between Dependent Variables in Study 2 (N = 101)
L.
DV
Napkins
Intention
Recall
Pre. SE
Post. SE
Control
Napkins
--
-.01
-.04
-.10
-.09
-.07
Intention
-.01
--
.24**
-.02
.56**
.64**
Recall
-.04
.24**
--
-.00
.26**
.16
L. Control
-.10
-.02
-.00
--
.16
.07
-.09
.56**
.26**
.19
--
.75**
-.07
.64**
.16
.07
.75**
--
Pre-treat
SE
Post-treat
SE
**Correlation is significant at p < .01.
75
Aside from pre-treatment self-efficacy, which was considered to be an
inherent individual difference in Studies 1 and 2, there were some additional
correlations among the dependent variables to note.
Behavioral intention was positively correlated with pre- and post-treatment
self-efficacy. Earlier studies have demonstrated that self-efficacy is a good predictor
of pro-environmental intentions (DeYoung, 1986) which explains the positive
correlation between intention and self-efficacy. The correlation for post-treatment
self-efficacy is higher than pre-treatment self-efficacy, suggesting that self-efficacy
becomes a better predictor of recycling intentions following the embodied
experience of cutting a tree down.
Recall is also positively correlated with behavioral intention, implying that
participants develop greater intention to recycle when they perceive less presence
during the embodied experience. This is an inconsistency that cannot be explained
with current data and should be explored further in future studies. However, it
should be noted that the correlation between recall and behavioral intention was
relatively low (r = .24) and may be a spurious finding.
Finally, pre- and post-treatment self-efficacy levels demonstrated a strong
positive relationship, replicating correlations from Study 1.
Discussion
These results yield important insights into the role of immersion and
agency of control in the process of embodied experiences. Level of immersion
directly influenced the realism that participants felt during the embodied experience
(i.e., presence, recall). Participants in the High Immersion condition felt
significantly higher levels of presence compared to those in the Low Immersion
condition and thus self reported more presence and also recalled less information
76
presented while they were engaged in the embodied experience.
Agency of control directly influenced the level of internal attributions (i.e.,
locus of control) participants made for the vicarious experience of tree-cutting.
Participants who had active control of their actions in the Self-Move condition
perceived greater control and made greater internal attributions for their actions
compared to those who embodied the tree-cutter and had passive control in the
Other-Move condition. This implies that having active haptic control in IVEs
induces greater sense of personal responsibility for that action.
Presence and recall differed significantly only when levels of immersion
were varied, and not by agency of control. This implies that increasing the overall
level of immersion through layers of sensorimotor stimuli is more effective in
creating a realistic embodied experience compared to providing layers of just motor
stimuli (i.e., haptic control).
Individual differences in perspective taking propensity directly influenced
recycling intentions and moderated the level of increase in actual behavior and selfefficacy. Participants with High Propensity self reported significantly higher
intentions to engage in pro-environmental behavior compared to Low Propensity
individuals, which replicates the result of actual napkin use in Study 1. But in Study
2, actual usage of napkins was influenced by agency of control. Low Propensity
participants used significantly less napkins when they experienced the Other-Move
condition, and High Propensity participants used significantly less napkins when
they experienced the Self-Move condition. It seems that when it comes to actual
pro-environmental behavior, Low Propensity individuals perform significantly
better when they are given guided movements. These results replicate the
previously discussed colorblind studies in that participants with low ability to
77
understand other minds or others’ experiences seem to be helped more significantly
by embodied experiences presented in IVEs compared to those with high abilities.
A possible explanation may be that without previously constructed
sensorimotor schema, Low Propensity individuals may find it more difficult than
High Propensity individuals to immediately construct a new schema on their own.
Consequently, they seem to benefit more from the guided movements in the OtherMove condition than High Propensity individuals. On the other hand, individuals
with High Propensity for perspective taking demonstrated greater proenvironmental behavior when they were given active haptic control. Results imply
that these individuals have higher intention to engage in pro-environmental
behavior compared to participants with Low Propensity, but they seem to require
embodied experiences with active haptic control in order for their intentions to
connect to actual pro-environmental behavior. It may be that having active haptic
control and actually making the motor movements on their own during the
embodied experience aids High Propensity individuals in constructing novel
sensorimotor schemata during the embodied experience.
Finally, all participants felt a significant increase in their self-efficacy in
terms of improving the quality of the environment following any embodied
experience, replicating findings from Study 1. But more specifically, participants
with Low Propensity for perspective taking demonstrated greater increases in selfefficacy after the embodied experience compared to participants with High
Propensities. This replicates findings from the previous colorblind studies in which
Low Propensity individuals demonstrated more favorable attitudinal changes than
those High Propensity individuals after embodied experiences in an IVE as a
colorblind person.
78
CHAPTER EIGHT: CONCLUSION
Summary of Findings
Study 1 revealed that embodied experiences in IVE and MS can both
significantly increase the belief that the self’s actions could improve the quality of
the environment. But this increase in self-efficacy was only connected to actual proenvironmental behavior when participants experienced the tree-cutting experience
through IVE. IVE elicited significantly higher perceptions of presence and control
compared to MS. Thus, the results of Study 1 suggested that presence and personal
control may be the processes through which embodied experiences influence selfefficacy and pro-environmental behavior.
Study 2 delved deeper into these underlying mechanisms by manipulating
the level of immersion and agency of control of the embodied experience. Levels of
immersion were manipulated by increasing the layers and the degree of vividness of
visual, aural, and haptic sensorimotor inputs provided to the participant. Higher
level of immersion resulted in higher self-reported presence and lower recall
compared to Lower level of immersion.
Agency of control was manipulated by allowing participants to either
actively control the haptic device on their own or to passively follow the
movements made by another participant. Participants who had active control in the
Self-Move condition made greater internal attributions for the damage they caused
to the environment (i.e., tree-cutting activity) compared to those who merely
followed guided movements recorded by another participant.
In both studies, any form of embodied experience, regardless of the level of
immersion or the agency of control, increased self-efficacy. But Study 2 revealed
that individual differences moderated the increase of self-efficacy with participants
79
with Low Propensity for perspective taking felt greater increases in self-efficacy
after the embodied experience than those with High Propensity. These results imply
that embodied experiences had a greater influence on Low Propensity individuals.
Individual differences also moderated recycling intentions and actual proenvironmental behavior. High Propensity individuals demonstrated higher
intentions to engage in pro-environmental behavior (Study 2) and actually engaged
in greater pro-environmental behavior (Study 1) compared to Low Propensity
individuals. However, High Propensity individuals engaged in significantly more
pro-environmental behavior when they had active haptic control of the motor
movements during the embodied experience. Conversely, Low Propensity
individuals needed guided haptic movement during the embodied experience to
actually engage in pro-environmental behavior.
Theoretical Implications – Embodied Cognition
Theorists have posited that virtual environments may be effective tools to
study the relationship between the body and the mind (Velmans, 1998), but few, if
any, studies have presented empirical support. The current dissertation studies are,
to the best of my knowledge, the first to demonstrate IVEs’ capacity to offer
embodied experiences that are sufficiently real to influence attitude and, more
importantly, actual behavior based on their virtual experience. My results confirmed
that embodiment through IVEs are more influential than MS (Study 1). Despite the
vivid descriptions of the tree-cutting process provided in the MS condition, actual
sensorimotor experiences during embodiment in IVE were more influential in
encouraging behavioral changes in the physical world. Results also confirmed that
the behavioral change may be moderated by individual differences in the propensity
for perspective taking and the agency of control (Study 2).
80
These results demonstrated that the effect of feeling vivid sensory cues and
making actual motor movements in IVEs are able to transfer into the physical world
to change actual behavior. Considering that participants were exposed to the
embodied experience within IVEs for about three minutes, the fact that this limited
exposure was able to decrease napkin usage by approximately 20 percent in both
studies is striking. Open-ended responses asking participants to guess the purpose of
the experiment ensured that none of the participants included in the final dataset
were consciously aware of using less napkins.
This is certainly not a result of modeled behavior as participants were not
learning how to cut down a tree through the embodied experience. Embodying the
novel experience of tree-cutting in the IVE after being primed with environmental
issues seems to have triggered or constructed relevant pro-environmental mental
schemata to later influence attitude and behavior in the physical world. The theory
of embodied cognition stipulates that the experience, perception, and retrieval of
embodied stimuli are closely linked and highly overlapping. Niedenthal (2007)
notes that through the interconnection of neurons that were active during the
original experience (e.g., seeing and hearing walnuts being cracked), only partial
activation of relevant neurons (e.g., only hearing walnuts being cracked) can begin
a cascade to trigger the activation of the complete set of neurons (e.g., as if one has
both seen and heard walnuts being cracked).
Following this logic, it is possible that the vivid experience of tree-cutting
in the IVE may have been interconnected with the information on non-recycled
paper products and deforestation given to participants during construction of mental
schemata. Being faced with a situation where they had to use paper napkins (to
clean up the water) may have activated schemata constructed during the
81
experimental treatment on using paper products and deforestation as well as the
vivid experience of cutting down a tree, ultimately eliciting participants to
unconsciously use less napkins. Similarly, it is likely that the partial activation of
relevant neurons, such as hearing the chain saw or watching a tree fall, may have
elicited pro-environmental behavior.
IVEs will be able to significantly expand the research on embodied
cognition by liberating users from the physical constraints of the body in the virtual
world (e.g., homuncular lobster). One of the main claims of embodied cognition
scholars are that cognition is “situated” (Wilson, 2002). In other words, cognition is
not an abstract and hard-wired entity that stands exclusively on its own, but rather
occurs in dynamic interactions with the body and the surrounding environment.
Another claim is that our thoughts and feelings are bound by our body. For example,
humans think and feel differently from a puppy because of the biological frame that
encases our mind – using fingers to grip, walking on two feet, and using our tongue
to speak languages. The current dissertation demonstrated that these boundaries can
be infinitely expanded with embodied experiences in IVE where users can construct
sensorimotor schemata embodying any self-representation in any situational context.
For instance, virtual environments allow individual to embody virtual
representations of different physical stature (Yee & Bailenson, 2007), sex/gender
(Yee, Ducheneaut, Yao, & Nelson, 2011), nationality (Groom, Bailenson, & Nass,
2009), and those with physical disabilities (Ahn & Bailenson, 2010; Kalyanaraman,
Penn, Ivory, & Judge, 2010). Taking advantage of the flexibility that virtual
environments provide in terms of self representation and environment, users are
able to build sensorimotor schemata that will later influence perception and
behaviors without having to experience the situation in the real world.
82
At the same time, IVEs can provide realistically vivid sensorimotor cues
that make the embodied experience more tangible than mere imagination. This is a
crucial advantage of IVEs because the importance of actual experiences in forming
sensorimotor schemata has been demonstrated in a recent study by Flusberg and
Boroditsky (2011). By having participants physically experience rotating the
wooden block figures presented in Shepard and Metzler’s (1971) mental rotation
task, results revealed that if participants find the physical rotation task to be difficult,
they will take longer to mentally simulate themselves physically rotating the same
object. Conversely, when participants were asked to mentally simulate the objects
rotating on their own (rather than the participants moving the object), the difficulty
of the physical task had no influence on how long they took to complete the mental
simulation. Thus, it seems that sensorimotor experiences that involve the self are
encoded separately from imagining the same situation without the self as the actor.
This is similar to the activation of canonical neurons discussed earlier (Di
Pellegrino, Fadiga, Fogassi, Gallese, & Rizzolatti, 1992), that respond selectively to
three-dimensional objects based on their shape, size, and spatial orientation.
Using IVEs, participants have the flexibility of the situated context, but are
actively able to experience the tangible sensorimotor stimuli of seeing, hearing, and
feeling. MS is also able to yield extensive freedom in terms of flexibility of situated
context (i.e., imagination is also free from many restraints of the physical world).
However, results of the current dissertation studies indicate that IVEs’ capacity for
vivid sensorimotor experiences may be a critical difference between embodied
experiences in IVE and MS, leading to changes in pro-environmental self-efficacy
and actual behavior. Because the vivid sensorimotor experience in the IVE led
participants to feel as if they were actively playing a part in causing deforestation by
83
cutting a tree down in the IVE forest, they may have felt greater personal
responsibility for the damage they incurred to the forest compared to participants
who merely imagined cutting down a tree.
Another important theoretical implication from the results of the current
dissertation is that individual differences in the capacity to feel presence and
respond to embodied experiences may be overcome with IVE. The encoding and
retrieval of embodied experiences is moderated by individual differences in the
sense that aspects of the experience most salient and important for the individual are
selectively preserved for future retrieval (Barsalou, 2003). In the colorblind studies
that I conducted previously, individuals with low propensity for empathy felt greater
similarity and developed more favorable attitudes toward colorblind people
compared to those with high propensities when they were provided with visual aids
from embodied experiences within IVEs. Study 2 of the current dissertation
replicated this effect. When Low Propensity individuals received guided motor
movements from another participant, the embodied experience successfully
translated into pro-environmental behavior in the physical world. Although further
investigations are needed to determine the exact underlying process, these results
imply that if Low Propensity individuals have difficulty in forming a novel
sensorimotor schema on the spot during the embodied experience, providing them
with guided sensorimotor cues may help. In Study 2, the individual difference in
capacity between Low and High Propensity groups was virtually eliminated with
guidance, and Low Propensity individuals engaged in just as much proenvironmental behavior as High Propensity individuals.
Theoretical Implications – Immersion and Presence in Virtual Environments
The results also contribute to the body of research on virtual environments
84
by individual manipulating and analyzing the relationships between the functional
features of virtual environments (i.e., immersion, agency of control) and the
subjective responses to these functional features (i.e., presence, perception of
control). In addition, my findings on how individual differences moderate these
processes will provide much needed empirical evidence in this field.
Not many studies to date have compared IVEs to MS in terms of how
individuals perceive presence. This may be because scholars generally assume that
the rich sensorimotor inputs provided within IVEs would naturally be more realistic
compared to imagination. However, there is empirical evidence to the contrary
which confirms that imagination may be more effective in producing vivid
sensorimotor cues. For instance, a narrative description of less than 45 words has
been shown to be more influential in forming false memories compared to digitally
manipulated photos (Garry & Wade, 2005). Also, imagery based advertisements
actually lowered purchase intentions for individuals with low ability to generate
mental imageries (Petrova & Cialdini, 2005).
Results from the current dissertation confirmed that IVE elicits higher
perception of presence (Study 1), but that only certain elements of the virtual
environment influence presence perception (Study 2). Contrary to my hypothesis,
Study 2 revealed that adding layers of sensory input (i.e., immersion) significantly
increased presence, but not adding layers of motor input (i.e., agency of control). It
seems that increasing the sense of presence through vivid sensory inputs should
precede adding interactivity or motor inputs when constructing embodied
experiences within virtual environments. In fact, Study 2 results imply that having
active motor control during an embodied experience may not be beneficial for
everyone – only individuals with High Propensity for perspective taking benefited
85
from active motor control.
These results replicate those of Slater and colleagues (2010). They placed
participants in a similar IVE setup as the current dissertation study using a HMD
and participants received three layers of sensorimotor inputs – visual, behavioral
mimicry, and touch. Of the three, they discovered that visual immersion by way of
stereoscopic vision and first-person perspective clearly dominated as an explanatory
factor for the perception of presence. Taken together with the results of the current
dissertation, it seems that the perception of presence is elicited mainly by sensory
cues rather than motor cues, and that presence does not necessarily increase
unconditionally with more layers of sensorimotor input.
Theoretical Implications – Environmental Studies
Most empirical studies in the realm of environment research have resorted
to self-report measures of behavioral intentions rather than observe actual
behavioral change as a result of experimental treatment, following the tradition of
theory of reasoned action (Fishbein & Ajzen, 1975) which predicts a positive
correlation between behavioral intention and actual behavior. However, as earlier
studies and the current dissertation studies demonstrate, self-efficacy and intention
may not be effective predictors of actual pro-environmental behavior. This gap may
be from social desirability bias (Fisher, 1993) in which respondents choose to report
“desirable” or socially accepted behaviors in surveys even if they do not reflect
accurate intentions (Sherman, 1980; Malhotra, 1993). Perception of control may
also play a role in creating a gap between intention and behavior in that the
behavior must be under complete volitional control (Ajzen & Fishbein, 1980). That
is, the individual must be free to decide at will to perform or not perform the
behavior without external barriers.
86
In the current dissertation studies, participants were asked to voluntarily
help clean up spilled water using however many napkins they wished to use. By
choosing an unobtrusive method of measuring behavior, I was able to obtain
naturalistic responses outside of the experimental context. Also, because the water
was spilled without notice, participants did not have time to worry about the
socially desirable or correct answer, decreasing possible confounds of social
desirability or demand characteristics5. Open-ended responses which asked
participants to guess the purpose of the experiment confirmed that none of the
participants included in the final datasets for both studies in the current dissertation
were aware of the purpose of the behavioral measure. That is, none of the
participants were able to consciously connect the napkin usage to their embodied
experience in the IVE or the MS treatment. Thus, a strong theoretical contribution
that these studies make is that self-reported measures of self-efficacy and behavioral
intention are assessed and compared against actual observation of paper
conservation.
Alternative Theoretical Explanations
Despite increasing interest and expanding empirical evidence on embodied
cognition, scholars are still unsure of the underlying mechanisms of how exactly the
experience, perception, and retrieval of embodied experiences work (Pezzulo et al.,
2011). The current dissertation demonstrated that embodied experiences in IVEs
may be influenced by different levels of immersion and agency of control and
moderated by individual differences in the propensity to engage in perspective
taking. However, there is still much to be learned about the exact processes of
embodied experiences. Although the framework of embodied cognition provides
partial explanation for the connection between embodied experiences in IVEs and
87
the resultant changes in attitude and behavior in the physical world, it may be
worthwhile to consider other theories that may supplement the theory of embodied
cognition in terms of understanding processes and underlying mechanisms.
One such theory is Zillman’s theory of excitation-transfer (1983), which
purports that the excitation or arousal from one stimulus can amplify the response to
another stimulus. For instance, if an individual is aroused by watching a violent film,
the residual excitement from this experience would amplify aggressive behavior in
an event that immediately follows the originally arousing experience, compared to
an individual facing the same event who did not experience the arousing experience.
This theory may provide further insight as to why IVE was more effective than MS
in eliciting greater pro-environmental behavior. Based on this theory, vivid
sensorimotor cues provided during the tree-cutting experience in the IVE were
likely to have led to greater arousal compared to mere imagination. This excitation
may have transferred into the physical world, triggering participants in the IVE
condition to engage in greater pro-environmental behavior compared to those in the
MS condition (Study 1).
However, it should be noted that the pro-environmental behavioral measure
in the current studies was gauged by observing the inhibition of behavior – using
less paper napkins. In this sense, the observed behavior in the current dissertation
studies yielded results that are the complete opposite of what would be expected
based on the excitation-transfer theory. That is, the theory would anticipate that the
IVE condition would yield greater arousal due to vivid sensorimotor cues provided
during the embodied experience compared to the MS condition. Heightened arousal,
in turn, would lead to greater manifestation of behavior – more napkin usage.
However, the results from Study 1 demonstrated that embodied experiences in IVE
88
actually led to less napkin use compared to the MS condition. This result discounts
the possibility of excitation and arousal driving the observed pro-environmental
behavior.
Another alternative explanation for the behavioral differences between the
two conditions may be that embodied experiences within IVEs trigger different
cognitive responses compared to MS. According to Schachter and Singer’s two
factor theory of emotion (1962), individuals experience emotional states based on
how their minds label the physiological excitation. For instance, when participants
interacted with a person of the opposite-sex in a highly arousing situation (i.e., on a
suspension bridge), they were likely to label the physiological excitation of standing
on a bridge as sexual attraction to the opposite-sex individual and later perceived
that individual as more attractive compared to participants who were in less
precarious locations (i.e., on a low, sturdy bridge) (Dutton & Aron, 1974). The
Schachter-Singer theory proposes that when individuals have no clear causal
explanation about their state of arousal, they will label arousal in terms of available
cognitions. Following this logic, it is possible that participants who experienced the
IVE condition in the current dissertation felt higher arousal compared to those in the
MS condition, and attributed the arousal to the information on using non-recycled
paper products and deforestation that they were primed with at the beginning of the
experiment. Consequently, it is likely that these participants labeled their arousal as
guilt or personal responsibility as a result of contributing to the deforestation
problem.
Lang’s limited capacity model of mediated message processing (2000) also
makes predictions relevant to arousing content of mediated experiences. According
to this model, arousing mediated content would allot greater cognitive resources for
89
the storage of the given information, which means less resource would be available
for encoding. This means that the overall amount of information encoded may be
reduced, but that the information that is encoded will be well stored. Relevant
studies show that emotionally arousing topics or the occurrence of such arousing
events are recalled well, but the specific information contained in the message is
remembered less well (Christianson, Goodman, & Loftus, 1992; Heuer & Reisberg,
1992). Applied to the results of the current studies, Lang’s model may explain why
significant changes to actual behavior could be observed in participants who were in
the IVE condition with arousing stimuli, when their recall scores were actually
lower than participants in the MS condition. That is, because the MS condition was
less arousing, participants may have had greater cognitive resources for encoding
specific information but less for storage. On the other hand, participants in the IVE
condition may have encoded less specific information but had more cognitive
resources for storage, implying that the information that was encoded was stored
well enough to influence their sensorimotor schemas and future behaviors.
Future research exploring embodied experiences in IVEs should take these
alternative explanations into consideration as possible underlying mechanisms that
drive attitudinal and behavioral changes as a result of embodied experiences. In
depth investigation of affect, arousal, and cognitive activities in response to
embodied experiences will yield greater insight into how embodiment is processed
by individuals and why they are motivated to change the way that they think and
behave in the physical world as a result of their virtual experience.
Applied Implications
The finding that vivid sensorimotor inputs presented through IVEs are able
to aid in the embodiment of completely novel experiences opens up great potential
90
for applications. For instance, public service announcements promoting proenvironmental behavior could become significantly more persuasive when
presented as an embodied experience in virtual environments rather than relying on
traditional media. Considering that a child between the ages of 8 and 18 spends an
average of nearly 1.5 hours at a computer and 1.25 hours playing video games every
day (Kaiser Family Foundation, 2010), the content of the public service
announcement could easily be designed into a video game.
With the advent of games with higher immersive features such as the
Nintendo Wii or the Microsoft Kinect, IVEs are becoming increasingly ubiquitous
(Ahn, Fox, & Bailenson, in press). Because video games played in IVEs have been
found to be more influential in terms of perception of presence and influencing
behavior (Perksy & Blascovich, 2007; 2008) pro-environmental messages relayed
through IVEs are likely to be more persuasive than using traditional media.
Furthermore, based on the results from the previously conducted colorblind studies
and the current dissertation studies, IVEs seem to be powerful enough to influence
attitude and behavior in any other novel context of embodiment. Thus, public
service announcements may be designed for highly immersive video game
platforms to relay messages on a wide range of themes such as health behavior or
safety.
In addition, the results also imply that individual differences in the capacity
to feel presence can be important factors that moderate the effect of embodied
experiences. This implies that content developers can tailor the message to meet the
needs of their target audience. By collecting data on individual differences such as
perspective taking propensity, more effective content of embodiment may be
developed to subject the individual to maximal influence from the experience. For
91
instance, for people with low propensity for perspective taking, the embodied
experience would provide sensorimotor aids that guide individuals to desirable
behavior. On the other hand, people with high propensities would be given full
control of the sensorimotor experience to learn through the embodiment.
In the same way, embodied experiences in IVE could significantly help
those involved in training and education. Lindgren (2009) found that embodying an
expert’s visual point of view in a navigating and search task significantly improved
task performance. In the same way, individuals could train and learn either through
active or guided sensorimotor experiences presented within IVEs. More importantly,
the training session can be tailored to meet each student’s needs. For instance, if a
student needs guided rather than active experiences, the student can practice the
guided movements for as long as he or she feels comfortable and then move on to
active experiences. Also, as demonstrated in Study 2, the sensorimotor movements
of a particular individual can be accurately tracked in the virtual environment and
played back to other users. This implies that experts’ sensory perceptions and motor
movements may be tracked to the millimeter and millisecond and played back to
novices so that they may embody the feelings and movements of an expert. Because
experts perceive objects and events in their domain of expertise differently than
non-experts (Goodwin, 1994), embodying their sensorimotor experiences is likely
to be a rich learning experience for novices.
Limitations and Future Directions
There are several limitations to the current dissertations studies that need to
be further explored in future research. Perhaps the most salient limitation of the
study is the embodied experience itself. The rich sensorimotor information provided
in the current dissertation studies, the vivid vibrations of the chain saw used to cut
92
down the tree in particular, served as a powerful manipulation that led to significant
behavioral differences between conditions. At the same time, the vividness of the
IVE stimulus may be a limitation because of the lack of generalizability. That is,
this particular context of embodied experience was chosen for its novel nature, but
the novelty of the experience (i.e., seeing, hearing, feeling the chain saw) may be
driving the differences observed between the conditions. Although one of the
greatest merits of using IVEs is that novel or even fantastic experiences can be
provided with realistic and vivid sensorimotor information, future studies should
also explore embodied experiences with greater mundane realism (i.e., turning off
lights for energy conservation, using less water for water conservation) within IVE
to confirm whether attitudinal and behavioral differences between embodied
experiences in IVEs and MS can still be observed.
The lack of evidence in terms of a clear psychological mechanism for the
results is another limitation of the current studies. For one, although both of the
studies imply that a new sensorimotor schema was constructed after the novel
embodied experience of cutting a tree down, I have no explicit evidence of this
internal representation. Future studies that involve assessment of mental maps that
individuals rely on during their decision making process may be effective. For
instance, Bagozzi and Dabholkar (1994) attempted to create a ‘goal map,’ looking at
how obtaining different goals may direct individuals’ recycling behaviors. In
addition, fMRI scans during the embodied experience may yield neurological
confirmation of this process and would make significant contributions to the
embodied cognition research.
Perception of presence and control were measured and analyzed in an
attempt to gain further insight into the underlying mechanisms of embodied
93
experiences. However, despite significant differences between experimental
conditions, there was a lack of positive correlations between these variables and
pro-environmental behavior, leaving room for speculation with regard to the exact
mechanism of embodied experiences. Because immersion and interactivity are all
affordances of a relatively novel medium – virtual environments – definitions and
measures are still in the processes of validation. This may explain the inconsistency
between measurements of recall and presence in Study 1 and 2. Future studies
should use both self-report and objective measures to further investigate the role of
presence and control during embodied experiences in virtual environments.
There were also some limitations in the measurements of self-efficacy and
actual pro-environmental behavior. First of all, although pre-treatment and posttreatment self-efficacy were measured and successfully demonstrated significant
differences between the two time periods, it is possible that post-treatment selfefficacy levels may have been influenced to some extent by the pro-environmental
information provided to the participants before experimental treatments rather than
the actual treatments (i.e., IVE and MS) themselves. It should also be noted that
different brands of napkins were used for Study 1 and Study 2, resulting in
differences in the absolute number of napkins used by participants to clean up the
water. However, the ratio of difference between conditions in napkin usage
remained similar in both studies. Finally, to measure behavioral change in terms of
napkin conservation rather than behavioral difference between conditions, a true
control condition is needed to assess the baseline of napkin usage. Similarly, a
condition that assesses the effect of actual tree-cutting activity in the physical world
on pro-environmental self-efficacy and behavior may yield greater insights to how
embodied experiences in IVEs compare against actual experiences in the real world.
94
Individual differences in perceiving presence is also an area of research that
lacks empirical evidence, and although these dissertation studies provide a good
foundation for initial investigation there is much room for improvement. Due to the
lack of earlier evidence on individual differences, I chose to look at group mean
differences based on a median split to factor individual differences into my analyses.
However, the method of applying a median split on a continuous variable has been
criticized (Aiken & West, 1991) for the loss of power. Thus, it is possible that there
may be further differences between individuals based on their disposition for
perspective taking that these dissertation studies were not able to assess.
In relation, although Study 2 confirmed significant interactions between
individual differences in the propensity for perspective taking and agency of control
for napkin usage, my hypotheses which expected higher levels of immersion and
self-based agency of control to elicit greater pro-environmental behavior compared
to lower levels of immersion and other-based agency of control were not supported.
There may be several reasons for this unexpected result. Although the current
analyses controlled for individual variances such as prior experience with
videogames or baseline levels of self-efficacy, there may have been an important
individual difference that was overlooked. It may also be that a more precise
operationalization of immersion is needed, as the current dissertation studies
incorporated a mixture of several sensorimotor cues in the operationalization of
high and low levels of immersion.
Finally, although increase in pro-environmental behavior after embodied
experiences was demonstrated in both studies, this behavior was measured only
after a short break (20-30 minutes) following the experimental treatment. Future
studies should explore longer lasting effects of embodied experiences in virtual
95
environments with either a follow-up survey as the one used in the colorblind
studies or longitudinal designs. Researchers should also look at the generalizability
of the embodied experiences after individuals have been taken out of the virtual
world. For instance, the behavioral measure assessed in the dissertation studies was
directly related to paper products, but it would be interesting to find out whether the
embodied experience of cutting trees could generalize to a wider range of proenvironmental behavior such as energy conservation.
These dissertation studies make meaningful contributions to prior research
on embodied cognition and virtual environments, as well as demonstrate exciting
potential for wide-ranging applications of embodied experiences through virtual
environments. The results demonstrate that our perceptions and actions within
virtual worlds can directly affect our perceptions and actions in the physical world.
With the rapid advancement and dispersion of technology, we may soon be in a
world in which we are able to embody any entity to virtually experience the life of
that entity augmented with vivid sensorimotor inputs, right in the comforts of our
own living room. If these virtual expeditions can lead to attitudinal and behavioral
changes in the physical world, it is imperative for scholars to conduct timely
investigations on the different effects of embodied experiences as this may easily
become a social issue. Through the current dissertation, I strived to take the initial
steps toward shedding light on these unanswered yet critical questions on the
increasingly blurred boundary between virtual and physical experiences.
96
NOTES
1. Entering race/ethnicity as a factor did not affect any of the subsequent analyses
and will not be discussed further.
2. Approximately equal numbers of participants from each group were randomly
assigned into IVE (High Propensity n = 10; Low Propensity n = 13) and MS
conditions (High Propensity n = 15; Low Propensity n = 8), with the exception of
one cell. However, results revealed no significant interactions between experimental
condition and perspective taking propensity and this issue is not discussed further.
3. In some of the studies discussed earlier, researchers looked at locus on control as
an inherent individual difference, but for the current study it is a state influenced by
the manipulation of agency of control rather than a trait and will be measured and
analyzed as such.
4. Open-ended responses assessed participants’ suspicion of the behavioral measure.
One female and one male were dropped from the final dataset due to suspicion.
5. The experimenter was visibly pregnant at the time. Consequently, none of the
participants were suspicious that the spilled water was a part of the experiment. This
was confirmed by open-ended responses taken after administering the behavioral
measure.
97
APPENDIX A: STIMULUS FOR MENTAL SIMULATION CONDITION IN
STUDY 1
You are in a forest with many tall, green, pine trees scattered around you. The
environment is placid and soothing, and it would be a nice place to have a picnic.
There are birds flying in and out of the trees, dancing in the skies, and you hear
them chirping. The birds fly past you, fly in front of you, and fly directly above
your head into the sky.
The forest floor is covered in rolling grassy patches and there are some hills to the
right side of you. Pretty, white clouds float above your head in the blue sky.
You hear the sound of running water and rustling leaves.
You are standing directly in front of a tall, pine tree with light brown colored bark.
It looks like a Christmas tree and the leaves along its branches that hang over your
head are spiky at the ends. The tree trunk is about the width of a person, maybe a
little bit more. The leaves and branches are so thick that they cover your view of the
sky.
To the left side of the tree, there is already a triangle shaped wedge cut out of the
trunk, and the tree looks like it is ready to be felled.
[BREAK. STOP READING AND WAIT FOR FURTHER INSTRUCTIONS]
Both of your hands are holding the handle of a chain saw with a warning sign which
reads that it is a dangerous tool. The saw is positioned at the right end of the tree
trunk, positioned to cut through until the saw reaches the part where a wedge has
already been cut out.
The chain saw starts with a roar and you feel the vibrations of the saw going into
the tree. It feels like cutting cardboard, almost. Very resistant and grainy. You feel
every slice of the saw and the vibrations from running the blade against the wood.
The grating noise from the chain saw increases and decreases as you pull the saw in
and out of the tree, slowly making your way through the trunk. The sound is so loud
and disruptive that it drowns out all the sounds from the nature. You can see
progress being made as you work and you see lighter bark inside the tree trunk
where you made the cut.
When your saw meets the part of the trunk where the wedge has already been cut
out, the trunk slow slides down to the left. You hear all the branches snapping and
cracking as they break, then a loud crash and great commotion as the trunk falls to
the ground, landing a couple feet away from the stump.
After the tree has fallen to one side, you suddenly see a large clearing in the forest.
It looks much more open and you can see all of the other trees that were once
hidden by this tall tree. It was a smooth, clean cut, and you see numerous rings on
the fallen tree trunk, realizing how old this tree was.
The forest is like dead silent now. There are no more birds flying around and their
98
chirping has stopped. It seems eerily empty- kind of in the middle of nowhere. Now
it feels more like a clearing instead of a thick forest.
99
APPENDIX B: STUDY 1 SURVEY ITEMS
Pretest
Perspective Taking Propensity
Please read the following statements carefully, and rate the extent to how well these
statements describe you (1 = Almost never; 2 = Once in a while; 3 = Sometimes; 4
= Frequently; 5 = Almost all the time).
1. How often do you attempt to understand your friends better by trying to figure
out what they are thinking?
2. How often do you try to think of more than one explanation for why someone
else acted as they did?
3. Overall, how often do you try to understand the point of view of other people?
4. When you are angry at someone, how often do you try to “put yourself in his or
her shoes”?
5. How often do you try to figure out what motivates others to behave as they do?
6. How often do you try to figure out what emotions people are feeling when you
meet them for the first time?
7. In general, how often do you try to understand how other people view the
situation?
Pre-treatment Self-Efficacy
Please read the following statements and indicate how well does each statement
describes your point of view (1 = Does not describe my point of view well; 2 =
Describes my point of view slightly well; 3 = Describes my point of view moderately
well; 4 = Describes my point of view pretty well; 5 = Describes my point of view
100
extremely well).
8. My individual actions would improve the quality of the environment if I were to
carpool instead of driving alone.
9. My individual actions would improve the quality of the environment if I were to
set my home appliances, such as the refrigerator, dishwasher, water heater, etc. to
“energy-saver” levels.
10. My individual actions would improve the quality of the environment if I were to
learn about the recycling facilities in my area.
11. My individual actions would improve the quality of the environment if I were to
attend a community meeting that involves concern over a local environmental issue.
12. My individual actions would improve the quality of the environment if I were to
buy resource conservation devices, such as low-flow faucet aerators for my sinks
and low-flow shower heads.
13. My individual actions would improve the quality of the environment if I were to
open windows for ventilation rather than using a fan or air conditioner.
14. My individual actions would improve the quality of the environment if I were to
buy products packaged in containers that either can be reused or recycled or are
made of recycled materials.
15. My individual actions would improve the quality of the environment if I were to
reduce the amount of my household trash by reusing or recycling items to the fullest
extent possible.
16. My individual actions would improve the quality of the environment if I were to
take my old tires to a recycling center.
101
17. My individual actions would improve the quality of the environment if I were to
buy and use recycled paper products.
Experimental Survey
Presence
(1 = Not at all; 2 = Slightly; 3 = Moderately; 4 = Pretty much; 5 = Very much)
1. To what degree did you identify with the tree-cutter?
2. To what extent do you feel that the tree-cutter is an extension of yourself?
3. To what extent do you feel that if something happens to the tree-cutter, it feels
like it is happening to you?
4. To what extent do you feel you embodied the tree-cutter?
5. To what extent did you feel you were in the same space as the tree-cutter?
6. To what extent did the tree-cutter seem real?
7. To what extent did you feel involved in the forest environment?
8. To what extent did you feel like you were inside a forest?
9. To what extent did you feel surrounded by the forest?
10. To what extent did it feel like you visited another place?
11. How much did the forest seem real?
Recall
12. Try to remember all the pieces of information on toilet paper production and its
effects on the forest. For instance, how many rolls of toilet paper is used in average
per year. We have prepared a form for you to use to record your
recollections. Simply write down the first information you remember in the first box,
the second in the second box, etc. Please put only one piece of information that you
102
remember in one box. IGNORE SPELLING, GRAMMAR, AND
PUNCTUATION. We have deliberately provided more space than we think most
people will need to insure that everyone would have plenty of room to write. So
don't worry if you don't fill every space. Just write down whatever information you
recall.
Personal Control
Rate your personal reaction based on your experience in the forest.
13. Looking at the picture above, enter the number under the drawing that best
describes how aroused you felt during your experience in the forest.
Post-treatment Self-Efficacy
14 – 23. Items 8-17 in the pretest.
103
APPENDIX C: SURVEY ITEMS FOR STUDY 2
Pretest
Items 1-17 from Study 1 Pretest
Experimental Survey
Control
1. Item 13 from Study 1 Experimental Survey.
Presence
Think back on your experience of standing in the shoes of the tree-cutter. Please
choose the adverb that best completes the sentence, based on your experiences (1 =
Not at all; 2 = Slightly; 3 = Moderately; 4 = Pretty much; 5 = Very much).
2. When standing in the shoes of the tree-cutter, I felt like my avatar was _______
an extension of my body within the virtual environment.
3. When standing in the shoes of the tree-cutter, I felt like my avatar was ______ a
part of my body.
4. When standing in the shoes of the tree-cutter, I felt ________like I could reach
into the virtual environment through my avatar.
5. When using my avatar, I felt_________like my arm was elongated into the
virtual environment through my avatar.
6. When standing in the shoes of the tree-cutter, I felt ______ like my own hand
was inside of the virtual environment.
7. When the tree-cutter's arms were moving in the forest, I felt ______ like I was
moving my own arms.
104
Think back on the experience you had in the forest, while you were standing in the
shoes of the tree-cutter. Please do not spend too much time on any one question.
Your first response is usually the best. For each question, choose the answer
CLOSEST to your own (1 = Strongly disagree; 2 = Disagree; 3 = Neither agree or
disagree; 4 agree; 5 = Strongly agree).
8. I felt that the tree branches and birds in the forest could almost touch me.
9. I felt I was actually visiting the forest.
10. I had the sensation that I moved in response to the movement of the chain saw.
11. I felt that I could almost smell the plant life in the forest.
12. I had a strong sense of sounds coming from different directions within the forest.
13. I felt surrounded by the forest.
14. I felt I could have reached out and touched the tree branches and the birds in the
forest.
15. I felt that all my senses were stimulated at the same time.
16. When cutting down the tree, I felt I changed the course of events in the forest.
17. I had the sensation that the tree, the birds, and the chain saw were responding to
me and my movements.
18. It felt realistic to move the chain saw in the forest.
Behavioral Intention
After you walk out of this room, please indicate the extent to which you plan to take
the following actions (1 = Rarely; 2 = Occasionally; 3 = Sometimes; 4 = Frequently;
5 = Usually).
105
19. How often will you purchase a product because it is packaged in reusable or
recyclable containers?
20. How often will you buy products made from recycled material?
21. How likely is it that you will switch from your current brand of toilet paper to
use a recycled brand (even if the recycled brand might be lesser in quality)?
22. How often will you watch TV programs to learn about the importance of
recycling paper?
23. How often will you talk to others about the importance of recycling paper?
Recall
24. Item 12 from Study 1 Experimental Survey.
Post-Treatment Self-Efficacy
25 – 34. Items 8-17 from Study 1 Pretest.
Locus of Control
Think back on your experience in the virtual forest while you took the perspective
of the tree-cutter. Think about what caused you (i.e., tree-cutter) to cut down this
tree. Then indicate the extent to which you agree with the following statements
regarding your reasons.
106
35. Is the cause(s) something... (1 = That reflects an aspect of yourself; 9 = That
reflects an aspect of the situation)
36. Is the cause(s) something... (1 = Inside of you; 9 = Outside of you)
37. Is the cause(s) something... (1 = Something about you; 9 = Something about
others)
107
APPENDIX D: STATISTICS FOR ALL ANALYSES IN STUDY 1
Napkins
(N = 46; IVE n = 23, MS n = 23, Low Propensity n = 21, High Propensity n = 25)
Main effects: Experimental condition, F(1, 40) = 4.16, p < .05, partial η2 = .09.
Perspective taking propensity, F(1, 40) = 4.73, p < .05, partial η2 = .11.
Interaction effect: Experimental condition x Perspective taking propensity, F(1, 40)
= .01, p > .05.
Self-reported Presence
(N = 46; IVE n = 23, MS n = 23, Low Propensity n = 21, High Propensity n = 25)
Main effects: Experimental condition, F(1, 40) = .21, p > .05.
Perspective taking propensity, F(1, 40) = 2.28, p > .05.
Interaction effect: Experimental condition x Perspective taking propensity, F(1, 40)
= .60, p > .05.
Recall
(N = 46; IVE n = 23, MS n = 23, Low Propensity n = 21, High Propensity n = 25)
Main effects: Experimental condition, F(1, 40) = 6.22, p < .05, partial η2 = .14.
Perspective taking propensity, F(1, 40) = 1.36, p > .05.
Interaction effect: Experimental condition x Perspective taking propensity, F(1, 40)
= .29, p > .05.
Control
(N = 46; IVE n = 23, MS n = 23, Low Propensity n = 21, High Propensity n = 25)
108
Main effects: Experimental condition, F(1, 40) = 4.44, p < .05, partial η2 = .10.
Perspective taking propensity, F(1, 40) = .58, p > .05.
Interaction effect: Experimental condition x Perspective taking propensity, F(1, 40)
= .14, p > .05.
Self-Efficacy
(N = 46; IVE n = 23, MS n = 23, Low Propensity n = 21, High Propensity n = 25)
Main effect: Self-efficacy, F(1, 41) = 4.62, p < .05, partial η2 = .10.
Interaction effects: Self-efficacy x Experimental condition, F(1, 41) = .16, p > .05.
Self-efficacy x Perspective taking propensity, F(1, 41) = 1.29, p > .05.
Self-efficacy x Experimental condition x Perspective taking propensity, F(1,
41) = 2.05, p > .05.
Control
(N = 46; IVE n = 23, MS n = 23, Low Propensity n = 21, High Propensity n = 25)
Main effects: Experimental condition, F(1, 40) = 4.44, p < .05, partial η2 = .10.
Perspective taking propensity, F(1, 40) = .58, p > .05.
Interaction effect: Experimental condition x Perspective taking propensity, F(1, 40)
= .14, p > .05.
109
APPENDIX E: MEANS AND STANDARD DEVIATIONS FOR ALL
DEPENDENT VARIABLES IN STUDY 2
Napkins
Immersion
Low
High
Total
Perspective Taking
Agency
Propensity
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Mean
8.0714
8.7273
8.3600
7.6667
6.9231
7.2800
7.8846
7.7500
7.8200
6.4615
9.1667
7.7600
9.4615
7.7692
8.6154
7.9615
8.4400
8.1961
7.2963
8.9565
8.0600
8.6000
7.3462
7.9608
7.9231
8.1020
8.0099
Std. Deviation
3.12470
2.28433
2.75197
3.28449
4.13242
3.69143
3.14104
3.46724
3.26821
3.38170
2.85509
3.36997
3.66550
3.19254
3.47651
3.77869
3.05614
3.41772
3.29119
2.54912
3.06001
3.53553
3.64354
3.61088
3.44051
3.24836
3.33315
N
14
11
25
12
13
25
26
24
50
13
12
25
13
13
26
26
25
51
27
23
50
25
26
51
52
49
101
110
Behavioral Intention - Total
Immersion
Low
High
Total
Perspective Taking
Agency
Propensity
Self-Move
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Mean
2.9429
2.9818
2.9600
2.4667
2.7846
2.6320
2.7231
2.8750
2.7960
2.4000
2.5167
2.4560
2.8308
2.6923
2.7615
2.6154
2.6080
2.6118
2.6815
2.7391
2.7080
2.6560
2.7385
2.6980
2.6692
2.7388
2.7030
Std. Deviation
.77828
.64159
.70711
.70496
.56250
.64208
.76958
.59509
.68867
.59442
.52886
.55534
.98267
.83513
.89625
.82544
.69637
.75727
.73643
.61919
.67879
.86317
.69919
.77704
.79200
.65600
.72642
N
14
11
25
12
13
25
26
24
50
13
12
25
13
13
26
26
25
51
27
23
50
25
26
51
52
49
101
111
Behavioral Intention – Item 1
Immersion
Low
High
Total
Perspective Taking
Agency
Propensity
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Mean
3.3571
3.6364
3.4800
3.2500
3.2308
3.2400
3.3077
3.4167
3.3600
3.1538
2.9167
3.0400
3.5385
3.2308
3.3846
3.3462
3.0800
3.2157
3.2593
3.2609
3.2600
3.4000
3.2308
3.3137
3.3269
3.2449
3.2871
Std. Deviation
1.08182
.92442
1.00499
1.05529
.83205
.92556
1.04954
.88055
.96384
.98710
.99620
.97809
1.19829
1.01274
1.09825
1.09334
.99666
1.04525
1.02254
1.00983
1.00631
1.11803
.90808
1.00976
1.06128
.94716
1.00336
N
14
11
25
12
13
25
26
24
50
13
12
25
13
13
26
26
25
51
27
23
50
25
26
51
52
49
101
112
Behavioral Intention – Item 2
Immersion
Low
High
Total
Perspective Taking
Agency
Propensity
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Mean
3.7143
3.8182
3.7600
3.1667
3.5385
3.3600
3.4615
3.6667
3.5600
3.0000
3.0000
3.0000
3.4615
3.3077
3.3846
3.2308
3.1600
3.1961
3.3704
3.3913
3.3800
3.3200
3.4231
3.3725
3.3462
3.4082
3.3762
Std. Deviation
.99449
1.07872
1.01160
.83485
.66023
.75719
.94787
.86811
.90711
1.08012
.85280
.95743
1.12660
.85485
.98293
1.10662
.85049
.98020
1.07946
1.03305
1.04764
.98826
.75753
.87088
1.02679
.88784
.95762
N
14
11
25
12
13
25
26
24
50
13
12
25
13
13
26
26
25
51
27
23
50
25
26
51
52
49
101
113
Behavioral Intention – Item 3
Immersion
Low
High
Total
Perspective Taking
Agency
Propensity
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Mean
3.2857
3.4545
3.3600
2.5000
3.0769
2.8000
2.9231
3.2500
3.0800
2.8462
2.4167
2.6400
2.9231
3.3846
3.1538
2.8846
2.9200
2.9020
3.0741
2.9130
3.0000
2.7200
3.2308
2.9804
2.9038
3.0816
2.9901
Std. Deviation
.99449
1.03573
.99499
1.00000
.95407
1.00000
1.05539
.98907
1.02698
.89872
.90034
.90738
1.38212
1.55662
1.46130
1.14287
1.35154
1.23701
.95780
1.08347
1.01015
1.20830
1.27460
1.25682
1.08934
1.18738
1.13574
N
14
11
25
12
13
25
26
24
50
13
12
25
13
13
26
26
25
51
27
23
50
25
26
51
52
49
101
114
Behavioral Intention – Item 4
Immersion
Low
High
Total
Perspective Taking
Agency
Propensity
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Mean
1.9286
1.6364
1.8000
1.5833
1.5385
1.5600
1.7692
1.5833
1.6800
1.2308
1.5833
1.4000
1.9231
1.4615
1.6923
1.5769
1.5200
1.5490
1.5926
1.6087
1.6000
1.7600
1.5000
1.6275
1.6731
1.5510
1.6139
Std. Deviation
.91687
1.20605
1.04083
.66856
.77625
.71181
.81524
.97431
.89077
.59914
.66856
.64550
1.11516
.66023
.92819
.94543
.65320
.80781
.84395
.94094
.88063
.92556
.70711
.82367
.87942
.81806
.84818
N
14
11
25
12
13
25
26
24
50
13
12
25
13
13
26
26
25
51
27
23
50
25
26
51
52
49
101
115
Behavioral Intention – Item 5
Immersion
Low
High
Total
Perspective Taking
Agency
Propensity
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Mean
2.4286
2.6364
2.5200
1.8333
2.5385
2.2000
2.1538
2.5833
2.3600
1.7692
2.6667
2.2000
2.4615
2.0769
2.2692
2.1154
2.3600
2.2353
2.1111
2.6522
2.3600
2.1600
2.3077
2.2353
2.1346
2.4694
2.2970
Std. Deviation
.85163
1.02691
.91833
1.02986
1.12660
1.11803
.96715
1.05981
1.02539
.59914
1.15470
1.00000
1.19829
.95407
1.07917
.99305
1.07548
1.03128
.80064
1.07063
.96384
1.14310
1.04954
1.08790
.97073
1.06266
1.02513
N
14
11
25
12
13
25
26
24
50
13
12
25
13
13
26
26
25
51
27
23
50
25
26
51
52
49
101
116
Recall
Immersion
Low
High
Total
Perspective Taking
Agency
Propensity
Self-Move
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Mean
3.7143
4.4091
4.0200
3.6250
3.5769
3.6000
3.6731
3.9583
3.8100
3.2308
3.0000
3.1200
2.5769
3.6923
3.1346
2.9038
3.3600
3.1275
3.4815
3.6739
3.5700
3.0800
3.6346
3.3627
3.2885
3.6531
3.4653
Std. Deviation
1.29666
2.13094
1.71075
1.49431
1.45554
1.44338
1.36340
1.80529
1.58078
1.54940
1.52256
1.50886
.95407
1.49358
1.35320
1.30399
1.51740
1.41719
1.41748
1.93419
1.65988
1.32822
1.44608
1.40385
1.37679
1.67458
1.53176
N
14
11
25
12
13
25
26
24
50
13
12
25
13
13
26
26
25
51
27
23
50
25
26
51
52
49
101
117
Pre- and Post-Treatment Self-Efficacy
Pre-treatment
Immersion
Low
Agency
Self-Move
self-efficacy
Other-Move
Total
High
Self-Move
Other-Move
Total
Total
Self-Move
Other-Move
Total
Post-treatment
Low
Self-Move
self-efficacy
Other-Move
Total
High
Self-Move
Other-Move
Total
Total
Self-Move
Other-Move
Total
Propensity
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Low Propensity
High Propensity
Total
Mean
2.9357
3.6818
3.2640
2.9167
3.3385
3.1360
2.9269
3.4958
3.2000
2.9385
3.4667
3.1920
3.0154
3.6077
3.3115
2.9769
3.5400
3.2529
2.9370
3.5696
3.2280
2.9680
3.4731
3.2255
2.9519
3.5184
3.2267
3.5571
3.7091
3.6240
3.5917
3.6923
3.6440
3.5731
3.7000
3.6340
3.3154
3.7667
3.5320
3.6385
3.9538
3.7962
3.4769
3.8640
3.6667
3.4407
3.7391
3.5780
3.6160
3.8231
3.7216
3.5250
3.7837
3.6505
SD
.67892
.31880
.65947
.80208
.41138
.65248
.72308
.40376
.65247
.95877
.58981
.83162
.92452
.72280
.86733
.92360
.65256
.84365
.80915
.48189
.74369
.85133
.59232
.76703
.82164
.53994
.75178
.86444
.58558
.74402
.90700
.73989
.80833
.86651
.65938
.76894
.86588
.67600
.79829
.92605
.74231
.83785
.89367
.70290
.82138
.85721
.62066
.76513
.89800
.73827
.81886
.87288
.67987
.79204
118
Locus of Control
Immersion
Low
High
Total
Perspective Taking
Agency
Propensity
Self-Move
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Self-Move
Low Propensity
High Propensity
Total
Other-Move Low Propensity
High Propensity
Total
Total
Low Propensity
High Propensity
Total
Mean
3.3810
3.1212
3.2667
2.5833
2.7433
2.6665
3.0128
2.9165
2.9666
3.3331
3.1667
3.2532
2.7436
2.7179
2.7308
3.0383
2.9333
2.9869
3.3579
3.1449
3.2599
2.6667
2.7306
2.6993
3.0256
2.9251
2.9768
Std. Deviation
1.67871
1.31886
1.50616
1.12927
1.20267
1.14645
1.48018
1.24415
1.35895
1.24677
1.06837
1.14349
1.33440
1.33226
1.30646
1.30046
1.20953
1.24524
1.45845
1.16700
1.32348
1.21716
1.24355
1.21875
1.37955
1.21378
1.29627
N
14
11
25
12
13
25
26
24
50
13
12
25
13
13
26
26
25
51
27
23
50
25
26
51
52
49
101
119
APPENDIX F: STATISTICS FOR ALL ANALYSES IN STUDY 2
Presence
Main effects: Immersion, F(1, 91) = 9.59, p < .01, partial η2 = .10.
Agency of control, F(1, 91) = .74, p > .05.
Perspective taking propensity, F(1, 91) = 0, p > .05.
Interaction effects: Immersion x Agency of control, F(1, 91) = .69, p > .05.
Immersion x Perspective taking propensity, F(1, 91) = .72, p > .05.
Agency of control x Perspective taking propensity, F(1, 91) = .65, p > .05.
Immersion x Agency of control x Perspective taking propensity, F(1, 91) =
0, p > .05.
Control
Main effects: Immersion, F(1, 91) = 1.37, p > .05.
Agency of control, F(1, 91) = 6.30, p < .05, partial η2 = .07.
Perspective taking propensity, F(1, 91) = .14, p > .05.
Interaction effects: Immersion x Agency of control, F(1, 91) = .41, p > .05.
Immersion x Perspective taking propensity, F(1, 91) = .10, p > .05.
Agency of control x Perspective taking propensity, F(1, 91) = .29, p > .05.
Immersion x Agency of control x Perspective taking propensity, F(1, 91) =
1.04, p > .05.
Behavioral Intention
Main effects: Immersion, F(1, 91) = 3.34, p > .05.
Agency of control, F(1, 91) = 0, p > .05.
120
Perspective taking propensity, F(1, 91) = 4.52, p < .05, partial η2 = .05.
Interaction effects: Immersion x Agency of control, F(1, 91) = 3.80, p > .05.
Immersion x Perspective taking propensity, F(1, 91) = .60, p > .05.
Agency of control x Perspective taking propensity, F(1, 91) = .17, p > .05.
Immersion x Agency of control x Perspective taking propensity, F(1, 91) =
2.84, p > .05.
Napkins
Main effects: Immersion, F(1, 91) = .11, p > .05.
Agency of control, F(1, 91) = .68, p > .05.
Perspective taking propensity, F(1, 91) = .49, p > .05.
Interaction effects: Immersion x Agency of control, F(1, 91) = 2.12, p > .05.
Immersion x Perspective taking propensity, F(1, 91) = .09, p > .05.
Agency of control x Perspective taking propensity, F(1, 91) = 4.86, p < .05,
partial η2 = .05.
Immersion x Agency of control x Perspective taking propensity, F(1, 91) =
1.59, p > .05.
Recall
Main effects: Immersion, F(1, 91) = 6.24, p < .05, partial η2 = .06.
Agency of control, F(1, 91) = .60, p > .05.
Perspective taking propensity, F(1, 91) = .10, p > .05.
Interaction effects: Immersion x Agency of control, F(1, 91) = .30, p > .05.
Immersion x Perspective taking propensity, F(1, 91) = .04, p > .05.
Agency of control x Perspective taking propensity, F(1, 91) = .41, p > .05.
121
Immersion x Agency of control x Perspective taking propensity, F(1, 91) =
2.47, p > .05.
Pre- and Post-treatment Self-Efficacy
Main effects: Self-efficacy, F(1, 92) = 41.93, p < .01, partial η2 = .31.
Interaction effects: Self-efficacy x Immersion, F(1, 92) = 0, p > .05.
Self-efficacy x Agency, F(1, 92) = 2.53, p > .05.
Self-efficacy x Perspective taking propensity, F(1, 92) = 8.91, p < .01,
partial η2 = .09.
Self-efficacy x Immersion x Agency of control, F(1, 92) = .04, p > .05.
Self-efficacy x Immersion x Perspective taking propensity, F(1, 92) = 1.79,
p > .05.
Self-efficacy x Agency x Perspective taking propensity, F(1, 92) = .03, p
> .05.
Self-efficacy x Immersion x Agency of control x Perspective taking
propensity, F(1, 92) = 1.21, p > .05.
Locus of Control
Main effects: Immersion, F(1, 91) = 0, p > .05.
Agency of control, F(1, 91) = 4.73, p < .05, partial η2 = .05.
Perspective taking propensity, F(1, 91) = 1.27, p > .05.
Interaction effects: Immersion x Agency of control, F(1, 91) = .02, p > .05.
Immersion x Perspective taking propensity, F(1, 91) = .01, p > .05.
Agency of control x Perspective taking propensity, F(1, 91) = .46, p > .05.
Immersion x Agency of control x Perspective taking propensity, F(1, 91)
122
= .22, p > .05.
123
REFERENCES
Ahn, S. J., & Bailenson, J. N. (2010). The effect of digitally augmented perspective
taking ability on motivation, empathic attitude, and helpful behavior.
Proceedings of the 60th Annual International Communication Association
Conference, June 22-26, Singapore.
Ahn, S. J., & Bailenson, J. N. (2011, in press). Self-endorsing versus otherendorsing in virtual environments: The effect on brand attitude and purchase
intention. Journal of Advertising.
Ahn, S. J., Fox, J., & Bailenson, J. N (2011, in press). Avatars. In Bainbridge, W. S.
(Ed.), Leadership in Science and Technology: A Reference Handbook. SAGE
Publications.
Aiken, L. S., & West, S. G. (1991). Multiple Regression: Testing and interpreting
interactions. Newbury Park, CA: Sage.
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social
behaviour. Englewood Cliffs, NJ: Prentice Hall.
Allen, J. B., & Ferrand, J. L. (1999). Environmental locus of control, sympathy, and
proenvironmental behavior: A test of Geller’s actively caring hypothesis.
Environment and Behavior, 31, 338-353.
Arns, L., Cook, D., Cruz-Neira, C. (1999). The benefits of statistical visualization in
an immersive environment. In Proceedings of IEEE Virtual Reality.
Bagozzi, R., & Dabholkar, P. (1994). Consumer recycling goals and their effect on
decisions to recycle: A means–end chain analysis. Psychology and Marketing,
11(4), 313–340.
Bailenson, J. N., Aharoni, E., Beall, A. C., Guadagno, R. E., Dimov, A., &
Blascovich, J. (2004). Comparing behavioral and self-report measures of
124
embodied agents' social presence in immersive virtual environments. Paper
presented at the 7th Annual International Workshop on PRESENCE,
Valencia, Spain.
Bailenson, J. N., Beall., A. C., Blascovich, J., Loomis, J., & Turk, M.
(2005). Transformed social interaction, augmented gaze, and social
influence in immersive virtual environments. Human Communication
Research, 31, 511-537.
Bailenson, J. N., & Yee, N. (2007). Virtual interpersonal touch and digital
chameleons. Journal of Nonverbal Behavior, 31, 225-242.
Bailenson, J. N., Yee, N., Blascovich, J., & Guadagno, R. E. (2008). Transformed
social interaction in mediated interpersonal communication. In Konijn, E.,
Tanis, M., Utz, S. & Linden, A. (Eds.), Mediated Interpersonal
Communication (pp. 77-99). Lawrence Erlbaum Associates.
Balderjahn, I. (1988). Personality variables and environmental attitudes as
predictors of ecologically responsible consumption patterns. Journal of
Business Research, 17, August, 51-56.
Bandura, A. (1994). Self-efficacy. In V. S. Ramachaudran (Ed.), Encyclopedia of
human behavior (Vol. 4, pp. 71-81). New York: Academic Press. (Reprinted in
H. Friedman [Ed.], Encyclopedia of mental health. San Diego: Academic
Press, 1998).
Barber, T.X., & Wilson, S.C. (1979). The Barber suggestibility scale and the
creative imagination scale: Experimental and clinical applications. The
American Journal of Clinical Hypnosis, 21(2-3), 84–108
Bargh, J. A., & McKenna, K. Y. A. (2004). The Internet and social life. Annual
Review of Psychology, 55, 573-590.
125
Barsalou, L. W. (1999). Perceptual symbol systems. Behavioral and Brain Sciences,
22, 577–609.
Barsalou, L. W. (2003). Situated simulation in the human conceptual system.
Language and Cognitive Processes, 18, 513-562.
Barsalou, L. W., Niedenthal, P. M., Barbey, A., & Ruppert, J. (2003). Social
embodiment. In B. Ross (Ed.), The Psychology of Learning and
Motivation, Vol. 43 (pp. 43-92). San Diego: Academic Press.
Beilock, S. L., & Holt, L. E. (2007). Embodied preference judgments: Can
likeability be driven by the motor system? Psychological Science, 18, 51-57.
Bernstein, E.M., & Putnam, F.W. (1986). Development, reliability, and validity of a
dissociation scale. Journal of Nervous & Mental Disease, 174, 727-735.
Biocca, F. (1997). The cyborg’s dilemma: Progressive embodiment in virtual
environments. Journal of Computer-Mediated Communication.
Biocca, F., Daugherty, T. M., Li, H., & Chae, Z. (2001). Effect of visual sensory
immersion on product knowledge, attitude toward the product and purchase
intention. In Proceedings of the Experiential E-Commerce Conference, F.
Biocca (Ed.). East Lansing, MI: Michigan State University.
Biocca, F., Harms, C., & Burgoon, J. K. (2003). Toward a more robust theory and
measure of social presence: Review and suggested criteria. Presence:
Teleoperators and Virtual Environments, 12, 456-480.
Biocca, F., Kim, J., & Choi, Y. (2001). Visual touch in virtual environments: An
exploratory study of presence, multimodal interfaces, and cross-modal
sensory illusions. Presence, 10(3), 247–265.
Blascovich, J., Loomis, J., Beall, A., Swinth, K., Hoyt, C., & Bailenson, J. N.
(2002). Immersive virtual environment technology as a methodological tool
126
for social psychology. Psychological Inquiry, 13, 103-124.
Boldero, J. (1995). The prediction of household recycling of newspapers – the role
of attitudes, intentions and situational factors. Journal of Applied Social
Psychology, 25, 440–462.
Bowman, N., Goodwin, J., Jones, P., & Weaver, N. (1998). Sustaining recycling:
Identification and application of limiting factors in kerbside recycling areas.
International Journal of Sustainable Development and World Ecology, 5,
263–276.
Boyd, C. (1997). Does immersion make a virtual environment more usable? In the
Proceedings of CHI ’97, Human Factors in Computer Systems: Looking to the
Future.
Bracken, C. C. (2005). Presence and image quality: The case of high-definition
television. Media Psychology, 7, 191-205.
Bradley, M. M., & Lang, P. J. (1994). Measuring emotion: The Self-Assessment
Manikin and the semantic differential. Journal of Behavioral Therapy and
Experimental Psychiatry, 25(1), 49-59.
Cacioppo, J. T., & Petty, R. E. (1981). Social psychological procedures for
cognitive response assessment: The thought-listing technique. In T. V.
Merluzzi, C. R., Glass, & M. Genest (Eds.), Cognitive Assessment (pp. 309342). New York: Guilford Press.
Campbell, R. (2000). Media and culture (2nd ed.). Boston: Bedford/St. Martin’s.
Christianson, S., Goodman, J., & Loftus, E. F. (1992). Eyewitness memory for
stressful events: Methodological quandaries and ethical dilemmas. In S.
Christianson (Ed.), The handbook of emotion and memory: Research and
theory (pp. 217–244). Hillsdale, NJ: Erlbaum.
127
Clark, A. (2001). Reasons, robots, and the extended mind. Mind & Language, 16(2),
121-145.
Cleveland, M., Kalamas, M., & Laroche, M. (2005). Shades of green: Linking
environmental locus of control and pro-environmental behaviors. Journal of
Consumer Marketing, 22(4), 198-212.
Cumming, J. (2008). Investigating the relationship between exercise imagery,
leisure time exercise behavior, and self-efficacy. Journal of Applied Sport
Psychology, 20, 184-198.
Curtis, F. A., Simpson-Housley, P., & Youck, C. (1984). Household energy
conservation and locus of control: A research note. Energy Research, 8, 89-93.
Davies, J., Foxall, G. R., & Pallister, J. (2002). Beyond the intention-behavior
mythology: An integrated model of recycling. Marketing Theory, 2(1), 29-113.
Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a
multidimensional approach. Journal of Personality and Social Psychology, 44,
113–126.
Decety, J., & Michel, F. (1989). Comparative analysis of actual and mental
movement times in two graphic tasks, Brain and Cognition, 11, 87-97.
Decety, J., Jeannerod M., & Prablanc, C. (1989). The timing of mentally
represented actions, Behavioral Brain Research, 34, 35-42.
DeYoung, R. (1986). Some psychological aspects of recycling: The structure of
conservation satisfaction. Environment and Behavior, 18, 435–449.
Di Pellegrino, G., Fadiga, L., Fogassi, L., Gallese, V., & Rizzolatti, G. (1992).
Understanding motor events: A neurophysiological study. Experimental Brain
Research, 91, 176–180.
Dutton, D. G., & Aron, A. P. (1974). Some evidence for heightened sexual attraction
128
under conditions of high anxiety. Journal of Personality and Social
Psychology, 30(4), 510-517.
Edwards, S. M., & Gangadharbatla, H. (2001). The effect of novelty on 3D product
presentations online. Proceedings of the Network Minds: Experiential 3D ECommerce, Michigan State University, September 27-29.
Ellis, S. R. (1995). Origins and elements of virtual environments. In W. Barfield, &
T. A. Furness III (Eds.), Virtual environments and advanced interface design
(pp. 14-57). New York: Oxford University Press.
Escalas, J. (2004). Imagine yourself in the product: Mental simulation, narrative
transportation, and persuasion. Journal of Advertising, 33(2), 37-48.
Feldman, J., & Narayanan, S. (2004). Embodied meaning in a neural theory of
language. Brain and Language 89, 385-392.
Feng, J., Spence, I., & Pratt, J. (2007). Playing an action video game reduces gender
differences in spatial cognition. Psychological Science, 18(10), 850-855.
Ferrari, P. F., Gallese, V., Rizzolatti, G., & Fogassi, L. (2003). Mirror neurons
responding to the observation of ingestive and communicative mouth actions
in the monkey ventral premotor cortex. The European Journal of
Neuroscience, 17, 1703-1714.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An
introduction to theory and research. Reading, MA: Addison-Wesley
Fisher, R. (1993). Social desirability bias and the validity of indirect questioning.
Journal of Consumer Research, 20, 303–315.
Flusberg, S. J., & Boroditsky, L. (2011). Are things that are hard to physically move
also hard to imagine moving? Psychonomic Bulletin & Review, 18(1), 158164.
129
Foroni, F., & Semin, G. R. (2009). Language that puts you in touch with your
bodily feelings: The multimodal responsiveness of affective expressions.
Psychological Science, 20(8), 974-980.
Fortin, D. R., & Dholakia, R. R. (2005). Interactivity and vividness effects on social
presence and involvement in a web-based advertisement. Journal of Business
Research, 58, 387-396.
Fox, J. (2010). The use of virtual self models to promote self-efficacy and physical
activity performance (Doctoral dissertation). Stanford University, Stanford,
CA.
Fox, J., Bailenson, J. N., & Binney, J. (2009). Virtual experiences, physical
behaviors: The effect of presence on imitation of an eating avatar.
PRESENCE: Teleoperators & Virtual Environments, 18(4), 294-303.
Gallagher, S. (2005). How the body shapes the mind. Oxford.
Gallese, V. (2000). The inner sense of action: Agency and motor representations.
Journal of Consciousness Studies, 7, 23–40.
Gallese, V., Fadiga, L., Fogassi, L., & Rizzolatti, G. (1996). Action recognition in
the premotor cortex. Brain, 119, 593-609.
Garry, M., & Wade, K. A. (2005). Actually, a picture is worth less than 45 words:
Narratives produce more false memories than photographs do. Psychonomic
Bullentin & Review, 12(2), 359-366.
Gehlbach, H., Brinkworth, M. E., & Wang, M.-T. (in press). The social perspective
taking process: What motivates individuals to take another’s perspective?
Teachers College Record.
Gerrig, R. J. (1993). Experiencing narrative worlds. New Haven, CT: Yale
University Press.
130
Glenberg, A. (1999). Why mental models must be embodied. In G. Rickheit & C.
Habel (Eds), Mental models in discourse processing and reasoning (pp. 7790). New York: Elsevier.
Goldman, A.I. (1989). Interpretation psychologized. Mind and Language, 4, 161–
185.
Goodwin, C. (1994). Professional vision. American Anthropologist, 96, 606-633.
Green, M. C., & Brock, T. C. (2000). The role of transportation in the
persuasiveness of public narratives. Journal of Personality and Social
Psychology, 79(5), 701-721.
Green, M. C., Brock, T. C., & Kaufman, G. F. (2004). Understanding media
enjoyment: The role of transportation into narrative worlds. Communication
Theory, 14(4), 311-327.
Green, M. C., Kass, S., Carrey, J., Herzig, B., Feeney, R., & Sabini, J. (2008).
Transportation across media: Repeated exposure to print and film. Media
Psychology, 11, 512-539.
Grezes, J., & Decety, J. (2001). Functional anatomy of execution, mental simulation,
observation, and verb generation of actions: A meta-analysis. Human Brain
Mapping, 1-19.
Grigorovici, D. M. (2003). Persuasive effects of presence in immersive virtual
environments. In G. Riva, D. Fabrizio, & W. Ijsselsteijn (Eds.), Being There:
Concepts, Effects and Measurement of Consumer Presence in Synthetic
Environments, (pp. 192-205). Ios Press, Amsterdam, The Netherlands.
Groom, V., Bailenson, J. N., & Nass, C. (2009). The influence of racial embodiment
on racial bias in immersive virtual environments. Social Influence, 4(1), 118.
131
Gruchalla, K. (2004). Immersive well-path editing: Investigating the added value of
immersion. In Proceedings of IEEE Virtual Reality.
Harlow, H. F. (1958). The nature of love. American Psychologist, 13, 673-685.
Hausenblas, H. A., Hall, C. R., Rodgers, W. M., & Munroe, K. J. (1999). Exercise
imagery: Its nature and measurement. Journal of Applied Sport Psychology,
11, 171-180.
Havas, D. A., Glenberg, A. M., Gutowski, K. A., Lucarelli, M. J., & Davidson, R. J.
(2010). Cosmetic use of Botulinum Toxin-A affects processing of emotional
language. Psychological Science, 21(7), 895-900.
Heeter, C. (1992). Being there: The subjective experience of presence. Presence:
Teleoperators and Virtual Environments, 1, 262-271.
Heeter, C. (2000). Interactivity in the context of designed experiences. Journal of
Interactive Advertising, 1 (1).
Heeter, C. (2003). Reflections on real presence by a virtual person. Presence, 12(4),
335-345.
Heineken, E., & Schulte, F. P. (2000). Acquiring distance knowledge in virtual
environments. Presented at the RTO HFM Workshop, April 13-15, Hague,
Netherlands.
Henion, K. E., & Wilson, W. H. (1976). The ecologically concerned consumer and
locus of control. In Henion, K.E. & Kinnear, T.C. (Eds.), Ecological
marketing, American Marketing Association Series (pp. 131-144). Austin, TX.
Heurer, F., & Reisberg, D. (1992). Emotion, arousal, and memory for detail. In S.
Christianson (Ed.), The handbook of emotion and memory: Research and
theory (pp. 151–180). Hillsdale, NJ: Erlbaum.
Jelfs, A., & Whitelock, D. (2001). Presence and the role of activity theory in
132
understanding: How students learn in virtual learning environments. Learning
Notes in Computer Science, 2117, 123-129.
Johnson, E. & Adamo-Villani (2010). A study of the effects of immersion on shortterm spatial memory. World Academy of Science, Engineering and
Technology, 71, 582-587.
Jostmann, N. B., Lakens, D., & Schubert, T. W. (2009). Weight as an embodiment
of importance. Psychological Science, 20(9), 1169-1174.
Kaiser Family Foundation (2010, January). Generation M2: Media in the lives of 8to 18-year-olds. Menlo Park, CA: Kaiser Family Foundation.
Kalyanaraman, S., Penn, D. L., Ivory, J. D., & Judge, A. (2010). The virtual
doppelganger: Effects of a virtual reality simulator on perceptions of
schizophrenia. The Journal of Nervous and Mental Disease, 198(6), 437-443.
Kim, T., & Biocca, F. (1997). Telepresence via television: Two dimensions of
telepresence may have different connections to memory and
persuasion. Journal of Computer-Mediated Communication, 3(2).
Kohler, E., Keysers, C., Umilta` , M. A., Fogassi, L., Gallese, V., & Rizzolatti, G.
(2002). Hearing sounds, understanding actions: Action representation in
mirror neurons. Science, 297, 846–848.
Lakoff, G., & Johnson, M. (1999). Philosophy in the flesh. The embodied mind and
its challenge to western thought. New York: Basic Books.
Landauer, T. K. (1962). Rate of implicit speech. Perceptual and Motor Skills, 15,
646.
Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for
categorical data. Biometrics, 33, 159-174.
Lang, A. (2000). The limited capacity model of mediated message processing.
133
Journal of Communication, 50(1), 46-70.
Lang, P. J. (1980). Behavioral treatment and bio-behavioral assessment: computer
applications. In J. B. Sidowski, J. H. Johnson, & T. A. Williams (Eds.),
Technology in mental health care delivery systems (pp. 119-137). Norwood,
NJ: Ablex.
Lanier, J. (2006). Homuncular flexibility. Edge.
Lee, K. M. (2004). Presence, explicated. Communication Theory, 14, 27-50.
Lee, K. M., Park, N., & Jin, S-A. (2006). Narrative and interactivity in computer
games. In P. Vorderer, & J. Bryant (Eds.), Playing computer games: Motives,
responses, and consequences (pp. 259-274). Mahwah, NJ: Lawrence Erlbaum
Associates.
Leiner, D. J., & Quiring, O. (2008). What interactivity means to the user: Essential
insights into and a scale for perceived interactivity. Journal of Computer
Mediated Communication, 14, 127-155
Lessiter, J., Freeman, J., Keogh, E., & Davidoff, J. (2001). A Cross-media presence
questionnaire: The ITC-Sense of Presence Inventory. Presence, 10(3), 282297.
Li, H., Daugherty, T., & Biocca, F. (2002). Impact of 3-D advertising on product
knowledge, brand attitude, and purchase intention. The mediating role of
presence. Journal of Advertising, 31(3), 43-57.
Lindgren, R. (2009). Perspective-based learning in virtual environments. (Doctoral
dissertation). Stanford University, Stanford, CA.
Lim, S., & Reeves, B. (2009). Being in the game: Effects of avatar choice and point
of view on psychophysiological responses during play. Media Psychology,
12(4), 348-370.
134
Lombard, M. (2001). Resources for the study of presence: Presence explication.
Available at: http://nimbus.temple.edu/~mlombard/Presence/explicat.htm
Lombard, M., & Ditton, T. (1997). At the heart of it all: The concept of presence.
Journal
Of Computer-Mediated Communication, 3(2), available at
http://jcmc.indiana.edu/vol3/issue2/lombard.html
Loomis, J. M. (1992). Presence and distal attribution: Phenomenology, determinants
and assessment. SPIE Vol. 1666: Human Vision, Visual Processing and
Digital Display III, 590–595.
Loomis, J. M., Blascovich, J., & Beall, A. C. (1999). Immersive virtual
environment technology as a basic research tool in psychology. Behavior
Research Methods, Instruments, & Computers, 31(4), 557-564.
Lozano, S. C., Hard, B. M., & Tversky, B. (2007). Putting action in perspective.
Cognition, 103, 480-490.
Malhotra, N. (1993) Marketing research – An applied orientation. London: Prentice
Hall.
Martin, K. A., & Hall, C. R. (1995). Using mental imagery to enhance intrinsic
motivation. Journal of Sport & Exercise Psychology, 17, 54-69.
McAuley, E., Duncan, T. E., & Russell, D. W. (1992). Measuring causal
attributions: The revised Casual Dimension Scale (CDSII). Personality and
Social Psychology Bulletin, 18(5), 566-573.
McCarty, J. A., & Shrum, L. J. (2001). The influence of individualism,
collectivisim, and locus of control on environmental beliefs and behavior.
Journal of Public Policy & Marketing, 20(1), 93-104.
Meehan, M., Razzaque, S., Whitton, M. C., & Brooks, F. P. Jr. (2003). Effect of
latency on presence in stressful virtual environments. In Proc. IEEE Virtual
135
Reality 2003 Conference, 22–26 March, Los Angeles, CA. 141–148.
Meinhold, J. I., & Malkus, A. J. (2005). Adolescent environmental behaviors. Can
knowledge, attitudes, and self-efficacy make a difference? Environment and
Behavior,37, 511-532.
Miles, L. K., Nind, L. K., & Macrae, C. N. (2010). Moving through time.
Psychological Science, 20(10), 1-2.
Minsky, M. (1980). Telepresence. OMNI Magazine.
Murray, C., Fox, J., & Pettifer, S. (2007). Absorption, dissociation, locus of control
and presence in virtual reality. Computers in Human Behavior, 23(3), 1347–
1354.
Murray, J. (1997). Hamlet on the Holodeck: The future of narrative in cyberspace.
Cambridge, MA: The MIT Press.
Nichols, S., Haldane, C., & Wilson, J. R. (2000). Measurement of presence and its
consequences in virtual environments. International Journal of HumanComputer Studies, 52, 471-491.
Niedenthal, P. M. (2007). Embodying emotion. Science, 316, 1002-1005.
Niedenthal, P. M., Halberstadt, J. B., & Innes-Ker, A. H. (1999). Emotional
response categorization. Psychological Review, 106(2), 337.
Nowak, K. L., & Biocca, F. (2003). The effect of the agency and anthropomorphism
on users’ sense of telepresence, copresence, and social presence in virtual
environments. PRESENCE: Teleoperators & Virtual Environments, 12, 481494
Parsons, L. M. (1994). Temporal and kinematic properties of motor behavior
reflected in mentally simulated action. Journal of Experimental Psychology:
Human Perception and Performance, 20, 709-730
136
Pausch, R., Proffitt, D., & Williams, G. (1997). Quantifying immersion in virtual
reality. Proceedings of the 24th SIGGRAPH, Annual Conference on
Computer Graphics and Interactive Techniques.
Persky, S., & Blascovich, J. (2007). Immersive virtual environments versus
traditional platforms: Effects of violent and nonviolent video game play.
Media Psychology, 10(1), 135-136.
Persky, S., & Blascovich, J. (2008). Immersive virtual video game play and
presence: Influences on aggressive feelings and behavior. Presence, 17(1), 5772.
Persky, S., & McBride, C. M. (2009). Immersive virtual environment technology: A
promising tool for future social and behavioral genomics research and
practice. Health Communication, 24(8), 677-682.
Petrova, P. K., & Cialdini, R. B. (2005). Fluency of consumption imagery and the
backfire effects of imagery appeals. Journal of Consumer Research, 32
(December), 442–452.
Pezzulo, G., Barsalou, L., Cangelosi, A., Fischer, M. H., McRae, K., & Spivey, M. J.
(2011). The mechanics of embodiment: A dialog on embodiment and
computational modeling. Frontiers in Psychology, 2(5), 1-21.
Rafaeli, S. (1988). Interactivity: From new media to communication. In R. P.
Hawkins, J. M. Wiemann, & S. Pingree (Eds.), Sage Annual Review of
Communication Research: Advancing Communication Science: Merging
Mass and Interpersonal Processes, 16, 110-134. Beverly Hills: Sage.
Raja, D., Bowman, D. A., Lucas, J., & North, C. (2005). Exploring the benefits of
immersion in abstract information visualization. In the Proceedings of HCI
International.
137
Ramus, C. A., & Killmer, A. B. C. (2007). Corporate greening through prosocial
extrarole behaviours – A conceptual framework for employee motivation.
Business Strategy and the Environment, 16, 554-570
Ratan, R. (2010). Self-presence, explicated. Paper presented at the 60th Annual
Conference of the International Communication Association, June 22-25,
Singapore.
Reeves, B., & Nass, C. (1996). The media equation. New York: Cambridge.
Rice, R. E., & Williams, F. (1984). Theories old and new: The study of new media.
In R. E. Rice (Ed.), The new media: Communication, research, and
technology (pp. 33-54). Beverly Hills, CA: Sage.
Roberts, D. J., Wolff, R., & Otto, O. (2005). The impact of display system and
embodiment on closely coupled collaboration between remote users. In R.
Schroeder & A. Axelsson (Eds.), Avatars at work and play: Collaboration d
interaction in shared virtual environments (pp. 131-149). Netherlands:
Springer.
Rotter, J. B. (1954). Social learning and clinical psychology. NY: Prentice-Hall.
Ruffman, T., Garnham, W., Import, A., & Connolly, D. (2001). Does eye gaze
indicate implicit knowledge of false belief? Charting transitions in knowledge.
Journal of Experimental Child Psychology, 80, 201-224.
Rushton, J. P., Fulkner, D. W., Neale, M. C., Nias, D. K. B. and Eysenck, H. J.
(1986) Altruism and aggression: the heritability of individual
differences. Journal of Personality and Social Psychology, 50, 1192–1198.
Sanchez-Vives, M. V., Spanlang, B., Frisoli, A., Bergamasco, M., & Slater, M.
(2010). Virtual hand illusion induced by visuomotor correlations. PLos ONE,
5(4).
138
Sas, C. (2004). Individual differences in virtual environments. In M. Bubak, G.
Dick van Albada, P. Sloot, & J.Dongarra (Eds.), Computational science –
ICCS 2004, fourth international conference, proceedings, Part III.Lecture
Notes in Computer Science (vol. 3038, pp. 1017–1024). Springer-Verlag.
Sas, C., & O’Hare, G. M. P. (2003). Presence equation: An investigation into
cognitive factors underlying presence. Presence: Teleoperators and Virtual
Environments, 12(5), 523–537.
Sas, C., O’Hare, G. M. P., & Reilly, R. (2004). Presence and task performance: An
approach in the light of cognitive style. Cognition, Technology & Work, 6,
53–56.
Schachter, S., & Singer, J. E. (1962). Cognitive, social, and physiological
determinants of emotional state. Psychological Review, 69(5), 379-399.
Shane, M. S., Stevens, M. C., Harenski, C. L. & Kiehl, K. A. (2009). Double
dissociation between perspective-taking and empathic-concern as predictors
of hemodynamic response to another’s mistakes. SCAN, 4, 111-118.
Shapiro, L. (2007). The embodied cognition research programme. Philosophy
Compass, 2(2), 338-346.
Sheehan, P. W. (1967). A shortened form of Betts’ questionnaire upon mental
imagery. Journal of Clinical Psychology, 23, 386–389
Sheppard, B. H., Hartwick, J., & Warshaw, P. R (1988). The theory of reasoned
action: A meta-analysis of past research with recommendations for
modifications and future research. Journal of Consumer Research, 15, 325–
343.
Sherman, S. J. (1980). On the self-erasing nature of errors of prediction. Journal of
Personality and Social Psychology, 39 (August), 211-221.
139
Shrum, L. J., Lowrey, T. M., & McCarty, J. A. (1994). Recycling as a marketing
problem: a framework for strategy development. Psychology & Marketing,
11(4), 393-416.
Singer, M. J., & Witmer, B. G. (1999). On selecting the right yardstick. Presence,
8(5).
Skalski, P., & Tamborini, R. (2007). The role of social presence in interactive
agent-based persuasion. Media Psychology, 10, 385-413.
Slater, M. (2004). How colorful was your day? Why questionnaires cannot assess
presence in virtual environments. Presence: Teleoperators and Virtual
Environments, 13, 484–493.
Slater, M., Alberto, C., & Usoh, M. (1995). In the building or through the window?
An experimental comparison of immersive and non-immersive walkthroughs.
Paper presented at VII Encontro Portugues de Computacao Grafica,
Eurographics. February, Monte de Caparica, Portugal.
Slater, M., Lanikis, V., Usoh, M., & Kooper, R. (1996). Immersion, presence, and
performance in virtual environments: An experiment with tri-dimensional
chess. In ACM Virtual Reality Software and Technology, VRST 1996, pp. 163172.
Slater, M., Spanlang, B., Sanchez-Vives, M. V., & Blanke, O. (2010). First person
experience of body transfer in virtual reality. PLos ONE, 5(5).
Slater, M., & Usoh, M. (1993). The influence of a virtual body on presence in
immersive virtual environments. In Proc. 3rd Annual Conference on Virtual
Reality (VR ‘93). 34–42.
Slater, M., Usoh, M., & Steed, A. (1994). Depth of presence in virtual environments.
Presence: Teleoperators and Virtual Environments, 3, 130-144.
140
Slater, M., & Wilbur, S. (1997). A framework for immersive virtual environments
(FIVE): Speculations on the role of presence in virtual environments.
Presence: Teleoperators and Virtual Environments, 6, 603-616.
Smith-Sebasto, N. J. & Fortner, R. W. (1994). The Environmental Action Internal
Control Index. The Journal of Environmental Education, 25(4), 23-29.
Speer, N. K., Reynolds, J. R., Swallow, K. M., & Zacks, J. M. (2009). Reading
stories activates neural representations of visual and motor experiences.
Psychological Science, 20(8), 989-999.
Stepper, S., & Strack, F. (1993). Proprioceptive determinants of emotional and
nonemotional feelings. Journal of Personality & Social Psychology, 64, 211220.
Steuer, J. (1995). Defining virtual reality: Dimensions determining telepresence. In
F. Biocca, & M. R. Levy (Eds.), Communication in the age of virtual reality
(pp. 33–56). Hillsdale, NJ: Lawrence Erlbaum Associates.
Strack, F., Martin, L. L., & Stepper, S. (1988). Inhibiting and facilitating conditions
of the human smile: A nonobtrusive test of the facial feedback hypothesis.
Journal of Personality & Social Psychology, 54, 768-777.
Sundar, S. S. (2000). Multimedia effects on processing and perception of online
news: A study of picture, audio, and video downloads. Journalism & Mass
Communication Quarterly, 77(3), 480-499.
Sundar, S. S., & Kim, J. (2005). Interactivity and persuasion: Influencing attitudes
with information and involvement. Journal of Interactive Advertising, 5(2), 629.
Sundar, S. S., Xu, Q., & Bellur, S. (2010). Designing interactivity in media
interfaces: A communications perspective. CHI 2010: Perspectives on
141
Design, April 10-15, Atlanta, Georgia.
Tabernero, C., & Hernandez, B. (2010). Self-efficacy and intrinsic motivation
guiding environmental behavior. Environment and Behavior, 20(10), 1-18.
Tellegen, A. (1982). Brief manual for the differential personality questionnaire.
Unpublished manuscript. Department of Psychology, University of
Minnesota, Minneapolis.
Thelen, E., & Smith, L. B. (1994). A dynamic systems approach to the development
of cognition and action. Cambridge, MA: MIT Press.
Thomas, L. E., & Lleras, A. (2009). Swinging into thought: Directed movement
guides insight in problem solving. Psychonomic Bulletin & Review, 16(4),
719-723.
Usoh, M., Arthur, K., Whitton, M. C., Bastos, R., Steed, A., Slater, M., & Brooks F.
P. Jr. (1999). Walking > walking-in-place > flying, in virtual environments.
In Proc. SIGGRAPH Computer Graphics Annual Conference Series. 359–
364.
Velmans, M. (1998). Physical, psychological, and virtual realities. In J. Wood (Ed.),
The Virtual Embodied (pp. 45-60). London: Routledge.
Vorderer, P. (2000). Interactive entertainment and beyond. In D. Zillmann & P.
Vorderer (Eds.), Media entertainment: The psychology of its appeal (pp. 2136). Mahwah, NJ: Erlbaum.
Wallach, H. S., Safir, M. P., & Samana, R. (2010). Personality variables and
presence, Virtual Reality, 14, 3-13.
Waller, D., Beall, A. C., & Loomis, J. M. (2004). Using virtual environments to
assess directional knowledge. Journal of Environmental Psychology, 24,
105-116.
142
Waterworth, J., & Waterwarth, E. L. (2006). Presence as a dimension of
communication: Context of use and the person. In G. Riva, M. T. Anguera,
B. K. Wiederhold, & F. Mantovani (Eds.), From communication to
presence: Cognition, emotions and culture towards the ultimate
communicative experience (pp. 81-96). Amsterdam: IOS Press.
Wells, G. L., & Petty, R. E. (1980). The effects of overt head movements on
persuasion: Compatibility and incompatibility of responses. Basic & Applied
Social Psychology, 1, 219-230.
Wenzel, E. M. (1992). Localization in virtual acoustic displays. Presence:
Teleoperators & Virtual Environments, 1, 80- 107
Wicker, B., Keysers, C., Plailly, J., Royet, J. P., Gallese, V., & Rizzolatti, G. (2003).
Both of us disgusted in my insula: The common neural basis of seeing and
feeling disgust. Neuron, 40, 655-664.
Wiederhold, B. K., Davis, R., & Wiederhold, M. D. (1998). The effects of
immersiveness on physiology. In G. Riva et al. (Eds.), Virtual environments
in clinical psychology and neuroscience (pp. 52–60). Amsterdam: IOS Press.
Willems, R. M., Hagoort, P., & Casasanto, D. (2009). Body-specific representations
of action verbs: Neural evidence from right- and left-handers. Psychological
Science, 20(10), 1-8.
Williams, L. E., & Bargh, J. A. (2008). Experiencing physical warmth promotes
interpersonal warmth. Science, 322, 606-607.
Wilson, M. (2002). Six views of embodied cognition. Psychonomic Bulletin &
Review, 9(4), 625-636.
Wirth, W., Hartmann, T., Bocking, S., Vorderer, P., Klimmt, C., Schramm, H., et al.
(2007). A process model of the formation of spatial presence experiences.
143
Media Psychology, 9(3), 493–525.
Witmer, B. G., & Singer, M. J. (1998). Measuring presence in virtual environments:
A presence questionnaire. Presence, 7(3), 225-240.
Yee, N., & Bailenson, J. N. (2006). Walk a mile in digital shoes: The impact of
embodied perspective-taking on the reduction of negative stereotyping in
immersive virtual environments. Proceedings of PRESENCE 2006: The 9th
Annual International Workshop on Presence. August 24-26, Cleveland, Ohio,
USA.
Yee, N. & Bailenson, J. N. (2007). The Proteus Effect: Self transformations in
virtual reality. Human Communication Research, 33, 271-290.
Yee, N., Bailenson, J. N., & Ducheneaut, N. (2009). The Proteus Effect:
Implications of transformed digital self-representation on online and offline
behavior. Communication Research, 36 (2), 285-312.
Yee, N., Ducheneaut, N., Yeo, M., & Nelson, L. (2011). Do men heal more when in
drag? Conflicting identity cues between user and avatar. Proceedings of CHI
2011, May 7-12. Vancouver, BC, Canada.
Zhong, C., Bohns, V. K., & Gino, F. (2010). Good lamps are the best police:
Darkness increases dishonesty and self-interested behavior. Psychological
Science, 21(3), 311-314.
Zhong, C., & Leonardelli, G. J. (2008). Does social exclusion literally feel cold?
Psychological Science, 19(9), 838-842.
Zillmann, D. (1983). Transfer of excitation in emotional behavior. In J. T. Cacioppo
& R. E. Petty (Eds.), Social psychophysiology: A sourcebook (pp. 215-240).
New York: Guilford Press.
144